Gendered Spaces of Informal Entrepreneurship in Soweto, South Africa

Abstract

The full diversity of economic activities in the informal economy is not well understood. The apartheid legacy of limiting informal entrepreneurship explains the most obvious geographical patterns in South Africa. After a decade and a half since the end of apartheid, however, knowledge about the places from which informal entrepreneurs operate is incomplete. Retailing is overemphasized and production firm entrepreneurs, both male and female, remain a neglected spatiality. This study reports on a survey of 100 firms in Soweto and in-depth interviews with 30 firm owners. The findings challenge the representations of isolated urban entrepreneurs dependent on inherited social capital and of women's exclusive engagement in retail. Entrepreneurs create their own social capital in work-related realms, but there are important gender differences. More finely tuned conceptualizations of entrepreneurs and of gendered working spaces need to be developed so that policy does not perpetuate unitary myths and incomplete spatial representations. [Key words: entrepreneurship, Soweto, informal economy, gender.]

Until lions learn to write, hunters will tell their history for them.

—Makura (2008, p. xii)

Low levels of entrepreneurship prevail in South Africa compared to other developing countries (Rogerson, 2001; Dierwechter, 2004; Hanson, 2009). The 2009 Global Entrepreneurship Monitor (GEM) report portrays a “deteriorating” environment with a “disturbingly low rate of new and established firm activity” (Bosma and Levie, 2009, p. 43). However, GEM claims are made from meager data: a random sample of 2,000 formal entrepreneurs across the entire country and a survey of 36 national formal economy experts. These data miss unregistered companies and are blind to place differences and within-place variations in what is obviously one of the most unequal economies in the world. For example, South Africa's townships (residential areas reserved for nonwhite populations during the apartheid era) are among the most economically marginalized components of cities. All too often, calls to assess the informal economy—one neighborhood at a time with greater gender sensitivity (Devey et al., 2003; Alderslade et al., 2006)—go unanswered, and changes that are taking place in South Africa's townships remain understudied.

There is a substantial international literature on women's work in the informal economy (Chen et al., 1999; Chen, 2001; Chant and Pedwell, 2008) as well as a sizeable body of scholarship on South African urban contexts (Lund and Skinner, 1999; Valodia, 2002; Dierwechter, 2004). In South Africa, the range of women's choices is typically understood to be extremely narrow in terms of sectoral participation. In the past, women's entrepreneurship was concentrated within retailing, especially home-based trading, handicraft market trading, and street trading (Todes, 1995; Ligthelm, 2007; Herrington et al., 2009). This retail dominance was reinforced by an official tilt in informal economic policy toward trading that has even found expression in township renewal policies that emphasize the development of retail centers. There are grounds for arguing that retailing in the informal economy as well as the concentration of women in this economic sector has been overemphasized at the expense of understanding the full diversity of informal economic activities.. Expert organizations such as South African Women Entrepreneurs' Network (SAWEN) warn against extrapolating from predefined and stereotypical retail portraits of women in informal economy (SAWEN, 2005). It needs to be underscored that the knowledge base on South African informal entrepreneurship requires continued enlargement so as to avoid overemphasizing a unitary construct as well as a static narrative on women's participation in the informal economy. Most importantly, research on nonretail firms in the South African context needs updating and enlarging so that more sophisticated and meaningful policies can be developed concerning diverse work sites in the informal economy, particularly sites in the townships (Rogerson, 1997; Rogerson and Rogerson, 1997; Chandra et al., 2001).

The emergence of new economic activity lies at the heart of entrepreneurship, but informal economic production is underconceptualized (Dierwechter, 2002; Williams, 2006; Obeng-Odoom, 2011) and extensive on-the-ground research is needed for its assessment (Grant, 2009; Nijman, 2009; Grant and Oteng-Ababio, 2012). Production encompasses both manufacturing and services but has wider economic relevance. Following Hanson (2009, p. 251), an entrepreneur is “an individual who owns a business, assumes business risk, deals with uncertainties, takes charge of day-to-day management, provides employment,” and produces goods and services, some of which are aimed (though not exclusively) at the informal economy. No agreed-upon definition of informal firms exists, but several dimensions of informality are fundamental to township business environments: low levels of firm registrations; major deficiencies in firm accounting practices; non-use of business contracts; and high propensities to integrate work and life. Despite considerable variability and diversity among informal firms, stereotyping is common. Particular spatial and representational depictions of informal firm failures and the dearth of extreme success stories (e.g., Rogerson, 2001; Makura, 2008), of the legendary iconic economic prowess of market queens, and of male entrepreneurs' spearheading female workforces in industrial clusters (Cornwall, 2007) may well be overspent representations. Forging a contemporary geographical assessment of informal production firms in Soweto township, this study investigates: (1) if there are gender differences in where informal firms locate, and (2) if gender matters at the scale where firms are articulated within a spatial economy (e.g., neighborhood, city, national, and international). It builds on geographic thinking (Peberdy and Rogerson, 2000; Dierwechter, 2002, 2004; Hanson, 2009) that posits entrepreneurship as a fundamentally local process whereby the spatial practices of entrepreneurs entail linkages from the microspace of operation to the macrospace of the region and beyond.

This analysis connects two different strands of research, one focused on the creation of social capital within the networks of informal firms and the other examining entrepreneurship, gender, and place. These bodies of literature have not yet been combined to study informal production activities. I explore the experiences and network decisions of entrepreneurs as a way to open up the conceptualization of production entrepreneurship in the informal economy as well for the understanding of the township milieu. The focus is on Soweto, the largest and presently the most economically dynamic township in South Africa. Given the weight of historical processes, it is still the case that informal businesses gravitate to less visible township locales and longstanding gender discrimination further hinders women's economic opportunities.

This article is organized into four sections. The first reviews entrepreneurship, informal economy networking, and place, as well as the social capital creation literatures. The second section discusses the Soweto study area and informal production firm data collection. The third details the main findings of the study (general characteristics, sectoral diversity, location, linkages, social capital, and township entrepreneurial milieu). The final section discusses the policy implications of the research findings.

ENTREPRENEURSHIP, INFORMAL ECONOMY NETWORKING, AND PLACE

The significance that social networks play in the structuring of entrepreneurial activities in the informal economy is hard to overstate. The levels of unconscious cooperation of formal market economic relations do not apply to the informal arena. Instead, informal entrepreneurs engage in multifaceted processes encompassing a wide array of agents and support mobilizations structured by broader socioeconomic relations (Lourenço-Lindell, 2002). Nevertheless, informal firms network to access resources, such as business information, labor, market knowledge (i.e., suppliers, competitors, and customers), and financial support. A classical finding in economic—geographic research is that firm establishment tends to be on home turf and that this locality becomes the key action space for networking and launching local and extra-local linkages (Nijkamp, 2003). Entrepreneurs' ability to create a productive business is premised on who they are, where they are located, and how they interact within space (Hovorka, 2005). Co-location with other firms is always an advantage but not a requirement because networking benefits can also be realized within a common socioeconomic context (Murphy, 2006). Networking does not, however, always result in linked individuals and firms, nor does it guarantee freedom from subordination (Lourenço-Lindell, 2002). Despite networking, firms can evolve as “islands of production, disconnected from one another and isolated from, and often disdainful of, informal sector firms” (Murphy, 2003, p. 1878).

Clearly, locational geometries affect informal firms' networking, but most studies concentrate on a particular kind of place, that is, on a formal industrial cluster (Hanson, 2009). Formal industrial clusters (designated and organized industrial estates) create significant interfirm networking opportunities compared to home-based enterprises (HBEs,) which by their very nature offer limited interfirm networking dependent on personal-business connectedness (Gordon and Nell, 2006). The immediate vicinity is important, whether it be a low-income residential area, a business district, or a peripheral business location (Rogerson and Rogerson, 1997; Nijman, 2009). The township milieu offers specific advantages for informal entrepreneurs (e.g., tight social networks, weaker enforcement of planning regulations, affordable space, skilled immigrant workers) but also some obvious constraints (distance from established urban markets, such as inner Johannesburg, low-income vicinities, infrastructure and built form deficiencies, perceived high crime, and a xenophobic environment for non–South African entrepreneurs; e.g., successful Somali businesses are frequent targets of attack).

Gender intersects with other axes of difference (location, immigration status, and so forth) in shaping South African informal production spatialities (Dierwechter, 2004; Hanson, 2009; Hanson and Blake, 2009). Women's networks exhibit more homophily and more kin-based interactions than men's (Hanson and Blake, 2009), but solidarity cannot be taken as a given (Cornwall, 2007). Some gender patterning of activities reflect restrictions on domestic and productive responsibilities (child and educational responsibilities, and fears of divorce and abandonment are ever-present in the Southern African context) (Downing and Daniels, 1992). Moreover, female-headed households (more than one-third of all households in Soweto) are not relieved from unpaid work within and around the household, but some female entrepreneurs do employ women house-helpers to free up some of their time (Muthwa, 1994). An emerging thesis, which borrows from spatial mobility and labor market studies (Hanson and Pratt, 1988) is that women locate their businesses close to home as a consequence of their spatial entrapment (Downing and Daniels, 1992). Such locational constraints can entail a double marginalization reality: a peripheral location within a peripheralized community. Furthermore, as newcomers to entrepreneurship (in South Africa), women encounter restrictions to their sectoral participation (Herrington et al., 2009). Women's entry into the market is confined typically to niches (e.g., clothing and handicrafts) left over after more powerful males have entrenched their market positions (Downing and Daniels, 1992). Women's greater spatial consignment leads them to develop more locally circumscribed networks than men's (Lyons and Snoxell, 2005). Indeed, the spatial confinement of women's entrepreneurship in African contexts may dissuade HBE graduation to an industrial facility if it is located at a distance from home and neighborhood bases (Steel, 1994).

SOCIAL CAPITAL CREATION

In African urban contexts, there is a sizeable geographical literature on social capital and entrepreneurial networks (e.g., Murphy, 2003; McDade and Spring, 2005; Meagher, 2007). It is well established that state support programs do not reach deeply into the informal economy (Rogerson, 2001) and that women receive limited government recognition and support (Lourenço-Lindell, 2002; Hovorka, 2005). Accumulated evidence shows that ethnic-based contacts pivot at the center of business networking (Peberdy and Rogerson, 2000; Oteng-Ababio, 2011).

A modest but emerging research frontier emphasizes that entrepreneurs operating in the informal economy purposely develop business networks (Anderson and Jack, 2002; Lyons and Snoxell, 2005; Grant, 2010). Entrepreneurs make conscious decisions about how to deploy their assets in managing their businesses as well as their lives (McDade and Malecki, 1997; Lyons and Snoxell, 2005). Because they operate from off-the-beaten-path locations (e.g., homes and nontraditional, nonzoned workplaces), building trust is created in place-specific contexts that depend on geographically proximate as well as distant resources. Trust in business cannot be taken for granted, and it is not constituted the same way in every economic sector (requirements of production in terms of inputs, transformation, and outputs entail more interactivity). Also, trust is not static but rather a product of historical and contemporary specificity: evidence is accumulating that social capital in townships is now more in flux than before the end of apartheid (Piazza-Georgi, 2002; Jooste, 2005; Grant, 2010). In rapidly changing urban environments, inherited networks of family, kinship, and ethnicity appear to be dwindling but we know little about their replacement or reconfiguration.

Whereas refinements to notions of work and employment in the informal economy have been addressed (Webster, 2005; Grant, 2010), the gendered spaces of entrepreneurial work constitute an underexplored frontier (Lourenço-Lindell, 2002; Herod et al., 2007; Hanson and Blake, 2009). To extend our thinking on working space, I draw on a literature on building trust in economic space (MacKinnon et al., 2002; Murphy, 2003, 2006) and use empirical research (as opposed to theorizing) to assess what connects firms in a particular place at a particular time. Building on this, I tease out ways that trust is embodied in real work settings. This enables a spatial rethinking about social capital—geographical accounts of which have been considered as “almost dead” (Holt, 2008, p. 22).

ENTREPRENEURSHIP AND INFORMAL ECONOMIC SPACE IN SOWETO

Townships are unique spatial environments. With a historical legacy in apartheid policies and their spatial planning ordinances, townships functioned as remote residential dormitory localities for people working in the city. As spatial economic planning products, they were designed to eliminate economic opportunities for nonwhites. Their origins as nongenerative economic spaces pushed informal businesses (e.g., carpentry, refurbishment, clothing manufacturing) under the radar (Rogerson and Beavon, 1982; Webster, 1984), and the loosening of apartheid policies led to their further dispersal. The stubborn cartography of apartheid limited the development of a local spatial economy and in fact perpetuated spatial-economy disconnection between “black” townships and the “white city.” Retailing was the predominant local economic activity, but other economic activities functioned out of sight (Webster, 1984). Apartheid cities belonged to white settlers and blacks were excluded. Most black women (outside of domestic workers) were further overtly excluded through restricted access to cities that required permissions, which in turn were tied to their relationship to their male partner holding temporary work permits.

In the 1980s, the South African government implemented a drive to move Soweto “backyard” production into new incubator industrial estates such as Emdeni, Orlando West, and Dobsonville (Rogerson and DaSilva, 1988). Efforts to transform informal firms into formal ones registered some successes in Pennyville Estate on the perimeter of Soweto, but these enterprises more typically recorded high failure rates. Low occupancy rates became the norm in most estates; as a result, estate managers relaxed tenancy requirements to encourage informal manufacturing. This compromise gained currency after the dismantling of racial restrictions on business ownership in 1994.

Despite Soweto being officially incorporated into the Greater Johannesburg Metropolitan Area in 2001, the township still exhibits limited linkages to the wider urban economy (City of Johannesburg, 2008). Researchers determined that Soweto was home to 1.1 million residents (Crankshaw et al., 2000, p. 846), but most reckon the actual population to be much larger. According to official estimates largely based on formal economic activities, Soweto contributes only 4% of the metropolises' total economic activity even though Sowetans constitute 43% of its population (City of Johannesburg, 2008, p. 1). The disorganization of economic space within the township, with no apparent economicurban hierarchy, is a significant impediment. For example, the City of Johannesburg (2005, Annex 3) identifies 173 business areas (a mix of industrial estates and small-scale ribbons of economic activity) in addition to numerous backyard manufacturing and workshops scattered throughout Soweto.

In 2003, a new policy emphasis was instituted concerning economic development and internal employment opportunities in townships (Endres and Schenker, 2007; Technical Support Facility to the Urban Renewal Programme, 2009), focusing especially on retailing in its early stages (Ligthelm, 2007). A range of projects, such as paving roads, planting 60,000 trees, and constructing cycle lanes, parks, and new recreation spaces, improved the esthetics, health, and safety of the general urban environment. These improvements are further enhanced by City of Johannesburg township economic programs; for instances, R200 million was earmarked for projects to stimulate the township economy in 2008 (City of Johannesburg, 2008, p. 39).

DATA COLLECTION ON SOWETO MICRO PRODUCTION FIRMS

This study is based on responses to 100 firm questionnaires administered in May and June of 2009. The survey was conducted with the assistance of Community Agency for Social Enquiry (CASE) using a survey team of 10 multilingual field assistants. We had no comprehensive national or local data sources on informal firms to draw from, so a significant effort was devoted to assembling a list of “invisible” production firms. The analysis subsequently divided the 100 firms into male- and female-owned companies (no firm in the sample was jointly owned by a man and a woman).

Compiling this list proceeded through four steps. Step one involved conducting 320 household surveys in six low-income neighborhoods in Soweto, whereby respondents were asked to identify the locations of off-the-beaten-path informal firms in their neighborhoods. Step two entailed mining all listings of business nodes with an informal firm presence that had been identified by the City of Johannesburg (City of Johannesburg, 2005, 2008). Step three involved creating a master list (based on the lists generated from the previous steps) and engaging in extensive site reconnaissance to verify and update the lists with CASE field assistants. In step four, an extensive listing of companies was created for the areas that contained informal economic production activities. Subsequently, the team randomly sampled 100 firms in the listing from three types of business locations: 33 firms from formal business estates (areas planned and zoned for the purpose of promoting industrial development with site services, a formal service entrance, and ample parking); 33 firms from informal business nodes (ad-hoc business areas that developed spontaneously without site services and an entrance that is shared among firms); and the remainder from HBEs (firms operating from home and/or small structures at the home that utilize family land, capital, labor, and often residential space). Figure 1 depicts a typical informal business node in Soweto.

Fig. 1. Informal business node, Soweto. Source: Author, 2009.

The population was randomly selected from the master list. A pilot test of four questionnaires was administered in April 2009. The final firm questionnaire contained 73 questions. We did not encounter any refusals, perhaps because we paid a modest sum to obtain a completed questionnaire. Paying respondents for their time has obvious drawbacks and meant that we had to vet the completed questionnaires returned by the field assistants. Nevertheless, we had ample time to examine questionnaires in detail, and in specific cases we eliminated some from the final survey pool. In total, we conducted 106 surveys but report on 100 completed and vetted. One can also attribute the high response rate to entrepreneurs' desire to learn more about the kinds of firms operating in their vicinities and to their sense of being overlooked in official reports on economic activity in Soweto.

Follow-up interviews were conducted with 30 firm owners in July 2009 (15 men, 15 women). These interviews took place with the assistance of a translator as needed. Interviewees were equally selected among HBEs, formal estates, and informal estates. Interviews were conducted opportunistically when interviewees wanted to tell their story. These 30 interviews focused on assessing the building of social capital in the business and on obtaining information about the locality's and township's informal business milieu. Whenever appropriate, this information was verified by interviewing industrial estate managers and City of Johannesburg officials at their Soweto offices.

RESEARCH FINDINGS

Workplace Locations

Limited production space for informal firms is apparent, reflecting the apartheid legacy that stifled township economic activity. Most entrepreneurs began operations on home turf, typically inside the house, in a backyard shack, or in a rented space nearby. Most firms in industrial estates even traced their beginnings to a home or a backyard shack. Rental spaces at business locations are sometimes shared by different start-ups, either by compartmentalizing the work space or by subdividing the work hours among different firms. Subdividing work spaces is very common among immigrant entrepreneurs.

The gendered geography of the composition of different business sites is striking (Fig. 2). Women-owned firms operate at diverse sites; production activities that are closely connected to retail (e.g., cutting, sewing, jewelry, pottery, food preparation) concentrate mostly in homes. Men-owned firms exhibit a dominant presence in formal estates and a concentration in the heaviest production activities (e.g., mechanical engineering, automotive repairs). The foothold of female entrepreneurs within formal business estates has not changed much in recent years: their firms are still limited to the smallest units. For example, in Pennyville Estate women's firms cluster in “the pens”—inside subdivided, small mesh units of factory space. Estate Managers in 2009 asserted that women's firms were shunning estates.

Fig. 2. Informal production activities by gender in Soweto. Source: Author's survey, 2009.

Instead, women show a strong preference for locating at informal business nodes (Fig. 3). Interviews with them reveal a process of informal networking that results in a spatial pattern of business concentration akin to that generated by chain migration. As with innercity entrepreneurs (Rogerson and Rogerson, 1997), township women entrepreneurs make contact with “pioneer” production firms to glean information before deciding on rental space. Many women note that formal rental inquiries in industrial estates go unanswered. The outcome of women's informal networking at informal business nodes creates a snowball effect that reinforces a spatial concentration of production firms within a particular building and adjacent buildings. Some women informants emphasized an advantage of informal business node locations as offering the opportunity to create more gender- sensitive business environments (i.e., safer as well as more comfortable for owners and female clienteles). Many prefer suppliers and customers not to be vetted by male security staff who serve as gatekeepers. Others, however, agonize about women's overconcentration. “Women's wars” among rival competitors that could swell into heightened hostility is a concern (Cornwall, 2007). An Orlando female entrepreneur recounted a “women's battle among clothing [firm] owners” as a byproduct of sexual networking with the same man: it spilled over into the working environment, and resulted in firms opting to disconnect with proximate firms. Indeed, on balance both women and men seemed to be aware of the precarious balance of gender composition in a given locality. However, women-owned firms outside of industrial estates were adamant against moving there, not because of distance from home but because of distance from their customer base. The composition of industrial areas (fearing marginalization in hidden, second-class units) and subordination to entrenched larger firms and suppliers were other major considerations. Male-owned firms, by contrast, hold high aspirations for formal industrial estate locations.

Fig. 3. Locations of informal production by gender (weighted). Source: Author's survey, 2009.

Even though important differences among business locales exist (in terms of rent, access, esthetics, signage, and control of the space), commonalities prevail in partitioning operating space. Two-room units are typical: a larger space utilized for production functions and sometimes for related storage, and a smaller space, whenever present, as a management office, customer sales point, and social meeting place). Entrepreneurs emphasize improvisation as well as sociability in day-to-day production activities. At times, estate managers' evince flexibility in their support of entrepreneurs by relaxing certain regulations and enforcement procedures as needed. Not only are rents reduced when necessary, but work space partitioning is tolerated to help cover rent and on occasion production activities in different sectors are allowed to occur on the same premises. For instance, one Orlando clothing firm fell on hard times, and its male owner's crisis response was to set up a catering business in the same work space. He sold food at lunchtime to neighboring estate firms as a stopgap measure until his primary business picked up. There are even examples of estates not acting against firms reallocating space so that migrant workers could be provided with sleeping quarters within the industrial space. Firms “invent” ways to accommodate other friendly firms and/or family businesses into their work space when times are hard.

Comingling of production operations and living is the norm in HBEs, although production can take place in a backyard structure, and the home may be used for business purposes, such as communications and receiving customers and suppliers). Men report that such arrangements are fraught with complications: space partitioning and spousal disagreements intertwine. For instance, the movement of furniture back and forth within rooms of houses is a conflict trigger, especially when the firm usurps furniture. Parceling of internal space is sensitive when it involves dual uses, such as utilizing children's desks for storage for business by day as well as for their intended purpose by night. The situation is less tense when the firm and household is headed by a single individual who does not have to negotiate with a spouse (women represent a large majority here), but this does not mean there are not disagreements with other members of the household about internal space organization. Women report similar tensions in other settings (Mirchandani, 1999), but married women in Soweto who are firm owners downplayed this tension compared to their male counterparts. Nonetheless, it was easy to observe HBE production operations surrounding and even engulfing living functions in a number of instances. For example, one firm operated a tannery wherein the inside bathroom facility is used for both hide curing and basic human needs; an adjacent garage was subdivided for tanning and pelt dressing, for industrial chemical storage, and for general household storage. This HBE further relied on the good will of neighbors to ensure that their production activities were not reported to municipal authorities.

Sectoral Niches

Considerable sectoral diversity is evident in informal production activities (Fig. 4). The largest numbers of firms participate in repairs and refurbishment (panel beating, upholstery, and door and window frames are common activities). Some machinery firms report engaging in repair activity on the side but switching back to manufacturing when customer demand increases. Construction (e.g., brickmaking, zozo [informal shacks] construction), coffin, furniture, clothing/fashion (traditional garments, shoes, jewelry, beading) manufacturing, and food (beyond self-employment, retailing) are other niche specializations.

Fig. 4. Informal production by sector. Source: Author's survey, 2009.

Metals, tombstones/coffins, machinery, and chemicals are production niches controlled exclusively by men. Construction and furniture manufacturing are also male-dominated activities. Women participate in a narrower range of subsectors that are more heavily concentrated in niches traditionally associated with women (e.g., fashion, food, and pottery). However, women are not economically engaged only in these sectors (Downing and Daniels, 1992): their production has expanded into printing, recycling, repair, and refurbishment. There is, however, modest evidence that several women participate in less dynamic product markets, such as bead-making and clothing firms, which record not only the slimmest product lines but the lowest profits. Production in these markets is more affected by narrow fashion trends, and static product ranges are indicative of survivalist strategies within industries. Firms with the largest profits are male-owned (e.g., tombstone/coffin manufacturing, auto repairs, printing, and machine production), firms that, not surprisingly, involve the largest investments in technology, workforce size, and industrial estate locations.

Informal Firms' Spatial Economy

Tight social networks generate word-of-mouth clienteles and help secure some credit, raw materials, cheap labor, and specialized services when needed at the lowest cost. The local dimension is also emphasized by some entrepreneurs linking with other local entrepreneurs in purchasing inputs (e.g., repair, refurbishment, and recycling firms operate at the most localized level in the township). However, the township market for sourcing inputs is still limited, and links to central Johannesburg predominate. The extensive dependence of Soweto on central Johannesburg means that vortex spatialities endure: pulling firms into central Johannesburg to purchase the majority of inputs (e.g., chemicals, hardware, food ingredients) from formal firms is part of a continuing apartheid legacy. Township renewal policies have not made much of a dent in displacing the old apartheid supply geography, even though the social geography has changed dramatically.

Despite being anchored in the township, firm output circulates beyond it. Firms link to local individual customers for half of all output (Fig. 5). Some are repeat customers, but most are not. An unreliable but somewhat improving informal circuit means that entrepreneurs engage in intense wormholing to get hold of a local customer base. Thirty-five percent of output is sold to Soweto's informal firms. A noteworthy finding is firms' incorporation within different arenas of the economy. For instance, 5% of output is purchased by government procurement programs. Examples of government-contracted work that connect with the township's informal operators include the Bane Pele program, which obtains school uniforms for the underprivileged from informal clothing manufacturers, and the City of Johannesburg's acquisition of banners and posters for Soweto-sponsored events from local informal printing firms.

Fig. 5. Social capital in business formation. Source: Author's survey, 2009.

Figure 6 shows the markets served by various production industries and the differences in market concentrations between female- and male-owned firms. This two-part figure illustrates the total market concentration of all production firms as well as overall market concentration. In some sectors (e.g., pottery), firm output is traded to markets at all spatial scales but in other sectors (e.g., chemicals) firms concentrate exclusively on the local vicinity. Overall, firms' outputs do involve multiple spatial economies: informal production is not entirely township-rooted (Dierwechter, 2002). Machinery, printing, and pottery entrepreneurs engage all arenas in the wider spatial economy. However, some metal-working firms operate as production extensions for larger firms, some of which are located outside Soweto, and in this regard they function as disguised wage labor (Nobanda, 1998) incorporated adversely into the urban economy. In contrast, some operations of auto repair firms are more positively incorporated. There are examples of such firms engaging in profitable subcontract work for larger northern-based firms. In these cases, the formal firm conceals the Soweto involvement: sending repair work out to township panel beaters would not be well received, given negative perceptions of unskilled township workers and repair work billing that does not reflect geographic labor cost differentials. Pottery and printing firms, although locally based, register the largest international orientation.

Fig. 6. Gendered concentrations of output of informal firms in the spatial economy. Source: Author's survey, 2009.

Gender difference in output orientation is striking, some of which reflects the different sectors into which men's and women's firms are incorporated. For instance, men dominate chemical- and metals-related work, auto repairs, and construction, generating intermediate products more likely to be sold across intermediate distances. By contrast, women-dominated sectors (e.g., fashion, pottery, and food) are the most locally rooted and trade largely within immediate vicinities; paradoxically, however, some of this output also trades over long distances, across borders in most cases.

Greater cross-border circuitry in women's production output activity can be explained by several factors. First, women's production activity may simply be more international; indeed, women in identical product categories register a greater presence than men in international markets (e.g., shoe manufacturing), although there is an example of a male-owned firm that plied the most distant market by designing shoes specifically for dance troupes in Nigeria. Second, immigrant entrepreneurs have a greater foothold in women-owned production firms; by contrast, immigrant male entrepreneurs have only a modest Soweto presence, limited to “zozo” construction (informal shacks) and repairs/refurbishment businesses. And third, the spatial entrapment of women might play an inherent role in overcoming the gendering of space. Interviews with female entrepreneurs highlight significant entry barriers (gender discrimination, stereotyping) in urban and national markets, so splitting attention toward the most local and the most nonlocal market is an obvious strategic approach. This may also help to further elucidate why women's firms appear less dynamic in terms of size, profitability, and technology.

Output for international markets, however, is determined as much by international customers' request as by owners' deliberate business strategy. I discovered a firm that refurbished engines for a Soweto customer base that, largely by happenstance, became integrated into an international repair circuit. It developed as follows: Soweto taxi drivers traveling between Soweto-Maputo and Johannesburg-Harare facilitate the transport of engines for repair across borders, connecting the Soweto firm with international firms (these engines are repaired with parts readily available in the Johannesburg marketplace); in time the international component of the operation accounted for the main production activity of the firm. An individual customer's request initiated the international dimension of the firm's activity, but it took an entrepreneur to assume the risk, seize an opportunity, deal with international (un)certainty, and hire additional workers so that the firm was able to service a network of taxi drivers and their international customers, thereby extending the spatial economy of the firm to an international entrepreneurial network.

Harnessing Initial Social Capital

Entrepreneurs are motivated to start a business by responding to the lack of employment opportunities for low-skilled workers in the formal economy (where many register former work experience) and by finding and seizing a market opportunity. Survivalism looms large, but there is emerging evidence that entrepreneurs are being positively propelled into entrepreneurship, offering some support of the township “normalizing” by its convergence toward more typical developing-country entrepreneurial contexts (Perry et al., 2007). One Emdeni woman categorized her venture into entrepreneurship, noting “there will not be work for us unless we create our own work.” There is now a greater diversity in entrepreneurial motivations and a gradual movement beyond enforced “black” entrepreneurship (Rogerson, 1997; Chandra et al., 2001).

Production entrepreneurs report heterogeneous social capital origins. Earlier studies showed a total reliance on friends, family, and neighbors (Chandra et al., 2001; Gordon and Nell, 2006). Today, traditional social support mechanisms no longer predominate: instead, respondents emphasize doing it alone and obtaining assistance from other people in the field and through business associations (Table 1). One Emdeni male entrepreneur put it this way, “I am in the do-it-yourself economy.… I liken the help I received to the manufacturing process itself, in that I mainly got the help of people who know” (as opposed to nonexpert assistance).

TABLE 1. Ties Among Informal Firms

	Owner gender
	Male	Female	Average
Friends, family, neighbors	15%	16%	15.0%
Business associates	20%	13%	15.0%
Other people in your field	25%	20%	22.5%
No help from anyone	40%	51%	47.2%
Source: Author's survey (2009).

Several factors explain the diversification of support mechanisms. First, prior employment experience (often in the formal economy) and serial entrepreneurship in the informal economy (entrepreneurs average 11 years of self-employment experience) create more opportunities to diversify away from traditional sources of social capital. Higher levels of perseverance and continuity, even if intermittent, is a function of greater costs sunk into production (e.g., 89% of firms own equipment, and 11% have secured bank loans to expand production and/or operations), so churning from one informal production activity to another is a serious undertaking. Second, greater attachment within a sector allows entrepreneurs to participate in business training programs and apprenticeships (formal and informal training): 17% valued this as the crucial support mechanism. Third, the more specialized knowledge base in manufacturing as opposed to that required for retailing activities elevates the role of other people in the field (experts) above the traditional support apparatus. The instrumental role of experts was strongly expressed by five entrepreneurs who “graduated” from street trade to production. And fourth, women more than men expressed the view that support now has to be anchored in the marketplace because an out-of-the-house focus loosens kin support. ¹

Nevertheless, support of friends, family, and neighbors is instrumental in certain cases. Some male entrepreneurs mention the transference of skills from a neighbor/family member (e.g., instruction on welding techniques). A small cohort of entrepreneurs report successes in maintaining overlapping network spaces, combining family and new business associates while forging a new firm's linkages. It is more commonplace for women to diversify from the family network. A number of women entrepreneurs kept business start-ups quiet from neighbors and family members for various reasons. For example, a female Pennyville clothing entrepreneur explained that, “improving my life, my Soweto, my community, and my children's future is like slow cooking” [as opposed to abrupt change], while also acknowledging that her efforts do not alter the larger gender-propriety power structures.

Gender differences in support ratings are evident. Women rely more on help from other people, mainly organizations and other female entrepreneurs, than men do. Interviews repeatedly reveal women's unequal financial access to credit. An Orlando HBE woman entrepreneur sums it up: “all credit goes to men in Soweto.” Women's greater reliance on business relationships may also be a function of associations that actively engage women. For example, the Sizanani Women's Resource Center provides training and mentoring to a number of women in pottery and fashion entrepreneurship. Indeed, women's sources of support reveal the greatest diversity. There are examples of situations where start-up equipment (e.g., sewing machines, industrial ovens, commercial refrigerators) are provided by churches and charity organizations (e.g., the Sparrow Ministries, Florida, South Africa) as well as examples of mentorship linking from the formal to the township informal economy (e.g., informal bed-and-breakfast [b&b] operators becoming “adoptees” of large hoteliers) (Business Day, 2010).

Creating Social Capital

Creating social capital for business is a core entrepreneurial task. Undoubtedly, social capital is created in many places, but interviewees claim they can identify which relationships are useful for business, which are not, and which are the trusted places. Entrepreneurs claim some successes in consciously creating islands of power (even though many acknowledge contacts do not always work out) against the weight of wider structural constraints. The average network size for Soweto entrepreneurs' inner circle is 10 business relationships, but some men claim 50 key business-based friendships. Neighboring firms, particularly those in the same production niche, are identified as the most pivotal contacts even though some are competitors in the same sector. Entrepreneurs express more trust toward neighboring firms (when firms are equivalent in size) than firms/individuals at a distance. However, linking to neighboring firms in the same sector in a locality is not a given: HBEs and firms in business clusters claim that isolation is often inevitable when larger neighboring firms dictate many of the conditions by which, and with whom, suppliers operate.

Entrepreneurs deliberately create, nourish, and adapt their networks as much as possible. Socializing with neighbors, attending church services, and participating in burial societies are typical activities with high participation levels and commitment, but no entrepreneur attributes business success to social ties alone. Social ties are useful for referrals, but a word-of-mouth clientele is often not good for the balance sheet: friends of friendsoriented business incur considerable time and effort, for example, in educating customers and collecting payment. These sentiments seem to support a withering of inherited social capital and their replacement with market-based ties. The pace of change toward more market-based ties is faster when firms cluster in a formal estate and slower in HBEs. Thus the process seems contingent on workplace milieu (Lyons and Snoxell, 2005).

For most entrepreneurs, especially women, investment in contacts is considered part of “work.” Women more frequently emphasize the time spent forging business ties, and it may be their minority entrepreneur status that accounts for their greater effort. The networking of entrepreneurs integrates workplaces and homes. Networks with workers, including those of neighboring firms, result in some entrepreneurs hiring workers from the same social network and even the same household. There are examples of engineering firms whose workers have to cooperate using heavy machinery when the procedure is especially intricate, and entrepreneurs report hiring from the same family (a father and son in one instance) to enhance workplace efficiency and cooperation. Firm owners highlight the advantages of locating near an informal settlement because of its reservoir of slightly more skilled and affordable human power. For example, panel-beaters emphasize the availability of more skilled Zimbabwean workers on an as-needed basis. Moreover, informal settlement workers are motivated to work and easy to recruit, avoiding many of the pitfalls associated with family and friends' labor. As such, there is an enlargement of entrepreneurial networks extending to workers' social networks, reaching into township informal settlements and beyond the confines of owners' neighborhoods and/or workplaces.

However, some entrepreneurs reluctantly engage their employees beyond work and worker referrals. Many are fearful of employees starting rival businesses, a concern that results in informal entrepreneurs' unwillingness to delegate and organize divisions of labor within the firm. Some entrepreneurs do it all by themselves as a survivalist reaction. A male entrepreneur from the Emdeni area of Soweto explained his total involvement in every aspect of his business: “only I know about the quality and materials of every job, and my buying of inputs has to be in line with what I charge.” In economic terms, only the entrepreneur knows the variable quality-to-price ratio of each production order.

Entrepreneurs identify meeting points as key nodes for conducting extra-work networking activities (for men, these are often drinking establishments and for women schools and markets). The home is typically a trusting place (but not for all women), and HBE operators express the lowest trust vis-à-vis the home-workplace. Male owners from industrial estates express the highest trust for home, for instance holding after-hours business meetings with labor brokers at home (where privacy is important so as not to send signals to existing workers). Given the tensions in the townships surrounding Somali, Bangladeshi, and other immigrant retail business owners (viewed as operating the most vibrant retail operations), production firm owners are careful to avoid a native backlash against immigrant workers so that immigrant workers are carefully incorporated into the workplace. This process usually commences with international workers most often recruited from home rather than in the workplace, dampening immediate xenophobic reactions from existing workers.

Although women are more often based in HBEs, they make extra efforts to develop a circuit of network points for business promotion, such as women's groups, clients' homes/ workplaces, schools, more established businesses, and the like. Gender differences in networking strategies can be explained, in part, by the different sectors of participation. Men are more involved in heavy machinery, autos, and metals, so clients are more likely to visit them and heavy machinery industries rely more on international skilled workers. Women engage in refurbishment and fashion, which require greater spatial mobility (even apart from standard household commitments), so networking off-site is incorporated. By and large, women's firms are smaller and their workforces almost exclusively rely on South African workers. The more local circuitry in women's movements provides support for the spatial entrapment thesis because their situations and processes of inclusion and exclusion facilitate engagement in certain sectors.

Women's and men's social capital creation differs in significant ways. Women's networking is split among neighboring firms, women's social groups, and neighborhood contacts. Their social groupings entail high levels of homophily (sewing groups, women's clubs, and societies). As one Orlando West interviewee put it, “as an outsider in business the only doors half-open for better and for worse are women's groups that can understand the situation of women in business.” Women highlighted their dependence on mutual personal knowledge (in the first instance) and support for each other's businesses by referring “good customers” whenever possible. It is more common for women-owned firms (than men's) to collaborate within cooperatives (where up to 10 companies, mostly near neighbors, combine to fill large orders). In these cases, women establish a separate formal company by contract but also maintain their informal firm (hedging both formal and informal firm options). Even though women's circuitry and social capital creation is neighborhood-based, the spatial dimensions of their networking reveal a jump in scale from neighborhood to international contexts (skipping the intermediate Johannesburg vicinity). In contrast, men's networking is far more extensive in intermediate space, reflecting prior work experience outside Soweto and efforts to maintain those contacts. Prevailing gender discrimination and women's lack of prior business experience in central Johannesburg and elsewhere (a feature of the apartheid era) explains the missing urban foundations upon which they can build.

SOWETO'S INFORMAL ENTREPRENEURSHIP MILIEU

An emerging challenge in entrepreneurship research is to move beyond an exclusive focus on profit-seeking behaviors of entrepreneurs toward a more meaningful comprehension of entrepreneurs' local roles (Wiklund et al., 2011). Informal economic entrepreneurship does not usually enrich owners nor operate at a total cost of society (decreased tax revenues and noncompliance of planning regulations are only part of the equation). Indeed, the microsocietal contributions of entrepreneurs in the informal sector might be more important than the small profits generated. Above all else, these entrepreneurs claim “to belong in Soweto,” a noteworthy change from the fragile sense of belonging in earlier times.

Importantly, in answering questions about their business futures, entrepreneurs responded with income, sales, and profits (traditional markers of success) but equally emphasized their role in uplifting the township milieu (nontraditional markers of success), such as self-empowerment, control of microspace, and contributions to sustaining the local economy rather than maintaining the status quo of leaking revenues and jobs to other areas. This finding is similar to what Hanson (2009) detailed in terms of women's entrepreneurship: specifically, that it is expressed as part of community well-being and as a distinctive relationship between business and place. Aspiring to contribute to uplifting localities may also be a risk-minimizing strategy to incorporate informal economic activities within the context of an improving township business environment. Indeed, in Soweto both men and women convey a highly localized community commitment, with women expressing it even more strongly. The legacy of excluding many women from the city under apartheid and women's current feelings of exclusion from urban Johannesburg networks is fostering their stronger sense of belonging and commitment to Soweto Township.

The principle of ubuntu—a person is a person through other people as well as a part of a collective—widely articulated in community studies in South Africa, and shapes a considerable part of entrepreneurs' networks of social relations in place, be they highly localized or extended across various scales (city, national, and international). Many take significant pride in the impact of their small actions to remake the township even though Soweto is shaped by larger processes, including historical patterns, current investment trends, and large formal economic entities. Entrepreneurs' strategic optimism contrasts with the predominant involution and hopelessness that characterized poor black (ethnic and family-based) entrepreneurship in earlier times (Chandra et al., 2001). No doubt this agency that entrepreneurs express must be understood within the context of apartheid norms, which were seen most vividly in highly unequal power relations surrounding work and non-home work sites. Also, such levels of strategic optimism must be tempered by an acknowledgment of the euphoria that prevailed around the FIFA 2010 World Cup tournament that permeated even informal business environments. No doubt the egos of informal production entrepreneurs is a central element (Williams, 2006). It is unknown if entrepreneurs have a heightened sense of their own agency in contributing to changing their immediate places of operation (compared to nonentrepreneurs), but I suspect they do.

CONCLUSIONS

Soweto's informal production entrepreneurs act in the “do it yourself economy,” embedding their firms within their township localities and linking beyond whenever possible. Entrepreneurs are hidden away in Soweto workshops, located in HBEs, industrial areas, and informal business nodes. Hence, the sum of their combined activities and efforts to shape the business milieu do not register on development radars. An overwhelming focus on current informal-economy policies on retailing, youths, and new entrepreneurs misses existing production talents and their local as well as wider contributions.

The evidence uncovered in Soweto in this research suggests a need to embolden the effective voice of township women in business toward multisector activities (and beyond narrow retail representations in South Africa). That women are launching businesses does not necessarily change their subordinate position within society. Yet, success is highly dependent on the relationship of the place in which they are embedded (Hanson, 2009), their business vicinity, and the articulation of the township within wider economic spaces. Gender matters in explaining the spatiality of firms in terms of location, sectoral niches, composition of entrepreneurial links, and their firms' integration into the wider spatial economies, findings that are not as clear cut in nontownship places (e.g., Ahl, 2003). Local economic development policies in townships need to start recognizing the more circumscribed activity spaces of women, their lower levels of representation in industrial clusters, and, perhaps by default, their greater participation within transnational networks.

Townships are unique milieus for informal economic activities. Life and work are integrated in various ways in informal entrepreneurship and in various workplaces, and social capital creation for business purposes needs to be considered in its own right. Importantly, there is a spatial integration of work and life in the township that is not possible in other informal-economy hubs (e.g., inner city). For too long now, urban and economic thinking has compartmentalized industrial districts and informal economic activities as separate arenas of inquiry rather than accept the heterogeneities of informal worlds and their working spaces. Perhaps the most striking finding of this study of Soweto production entrepreneurs is that so many multiple spatialites constitute the informal economy. So much for a single account of the informal economy in South African policy circles! Recognizing multiple worlds of work will only enrich theory, practice, and research in desperate need of broader understandings, and will make some progress toward achieving a democratic account of everyday life and work complexities in the informal economy.

Notes

¹Men and older women are generally disapproving; patriarchal notions endure that the “proper” role of a town- ship woman is as a hard-working domestic laborer and that women work for their husbands (Meintjes, 2001).