Theorising MSMEs in Contexts of Urban Violence

What are the relationships between and among small businesses, conflict, and peaceful development in contexts of urban violence? Here, the complex formal and informal divisions of economic, political and social power, authority, and legitimacy – and the many grey areas between legality and illegality, necessity and opportunism – create challenging conditions for business operations and for peacebuilding. A grounded understanding is required if peaceand development-positive interventions are to be successful. Yet both the peacebuilding and development potential of micro, small, and medium enterprises (MSMEs), and the specific dynamics of business and conflict in urban spaces, are underdeveloped in the literature assessing and promoting business potential to catalyse positive change. We therefore extract from a broad range of literature a typology representing the weight of the extant frameworks for understanding MSMEs in contexts of urban violence. We then use primary research to construct inductively a framework that captures how those living with urban violence themselves perceive businesses, their relationships to violence, and their impacts on in/security and under/development. Drawing out key areas of synthesis and tension, we propose directions for future study and practice related to small businesses in violent cities, emphasising the need to eschew simple understandings of actors, agency, and objectives in favour of a more nuanced and humane inquiry into the pain and potential inherent in the local context.


Introduction
Private sector actors have complex relationships with peace and conflict dynamics. Businesses have the potential to meaningfully foster peace and sustainable development (Miklian 2017;Ford 2015;Miller et al. 2019;Ganson & Wennman 2016;Oetzel et al. 2010;Peschka 2011) and are being asked to 'become full partners in broader peacemaking and peacebuilding assessment, planning and execution' (Ganson 2014: 71). At the same time, businesses can also undermine development and create or sustain conditions that foster conflict OHCHR 2008;Drohan 2004) and may face significant constraints on their agency (Ganson 2019a) with regard to conflict and peaceful development. These contradictions demand a better understanding of the presumed mechanisms of how and why businesses may motivate or not, and succeed or fail, in creating positive impacts for peace and development.
Two areas of note require further scrutiny. First is the role of smaller businesses -particularly local small and medium enterprises (SMEs). The literature typically links the development and peacebuilding potential of the private sector with large, predominantly Western multinationals operating in lower-and middleincome countries (Getz & Oetzel 2009;Banks 2016;Kanashiro & Starik 2016;Kolk & Lenfant 2012;Hoelscher & Rustad 2019). Yet this obscures the role of SMEs, and their markedly different abilities and incentives to engage in peace and development activities. When expanding further to consider micro, small, and medium enterprises (MSMEs) -which includes businesses such as sole traderships and those with up to 5 or 10 workers (OECD 2005) -we know even less about their ability to affect change, the constraints on their agency, or their motivations. For instance, capacity limitations, embeddedness in violent communities, and dependence on social networks -particularly those with illicit actors -may limit MSMEs peace or development potential (e.g., Hoffmann & Lange 2016). Conversely, due to their embeddedness, these businesses may be potentially more likely to play a role in conflict prevention activities (Felgenhauer 2007;Rettberg 2004;de Vidovgrad 2015) or be central players in supporting, for instance, post-conflict private sector development and disarmament, demobilisation, and reconciliation (DDR) activities (Strachan 2017).
These competing contentions make clear that the mechanisms underpinning the agency and peace and development potential of MSMEs are not fully understood. Though business for peace research is beginning to catch up to economics, development studies, business studies, conflict resolution, and entrepreneurship fields in looking at the nuanced and complex positionality of smaller and indigenous firms in fragile spaces (Barkemeyer & Miklian 2019;Katsos & AlKafaji 2019;Miklian & Bickel 2020) it 'has yet to identify if or how these businesses can meaningfully generate peace at any level in society' (Joseph et al. 2020: 4).
A second area is how businesses and conflict dynamics intersect in urban areas. As we approach 60% of global population being urban, cities are vitally important to the Sustainable Development Goal agenda, with Former UN Secretary General Ban Ki-Moon declaring that ' our struggle for global sustainability will be lost or won in cities' (Moon 2012). While cities offer profound opportunities, they also challenge development and peacebuilding trajectories due to organised criminal violence, civic conflict, urban political violence and warfare (Beall et al. 2013;Kilcullen 2015;Dilegge et al. 2019;Konaev 2019). Here, SMEs are often seen as key peace and development actors, being vital job creators and forming deep ties in their communities (World Bank 2018). Yet the evidentiary base for SMEs driving peace and development-positive changes is generally weak -and positive impacts anecdotal or presumptive -belying a reality where local businesses are deeply embedded in the complex dynamics of urban violence.
Broadly summarising, we might simplify the current terrain as being characterised by three overarching perspectives that we can interrogate: 1) A recognition that businesses potentially play a role in peace and development processes, but how or why (or why not) is not always clear or consistent; 2) A general under-recognition of how MSMEs engage to support or undermine peace and development in fragile contexts; and 3) A lack of focus on how the specifically urban dynamics of violence and conflict intersect with private sector engagement in peace and development processes.
In engaging with these perspectives, this article explores how businesses in contexts of fragility are characterised in literatures vis-à-vis their peace and development roles, and how this maps onto real world experiences of small businesses operating in violent cities. However, rather than taking an outside-in view that characterises how firms in violent contexts behave, or an instrumental view of what firms can or should do, we favour an inside-out perspective that recognises that business actors' embeddedness in contexts of violence should drive our understanding of MSMEs in violent cities. The article proceeds as follows. In Section 2 we briefly review the literature to extract a simplified typology that synthesises how businesses in contexts of urban violence are conceptualised across disciplines. Section 3 presents empirical findings from exploratory field work in the Langa community of Cape Town, South Africa, to identify and systematise how businesses are seen from the perspective of one set of MSME owners and urban stakeholders. In Section 4 we identify areas of conceptual intersection and tension between how businesses are conceptualised in the academic and policy literature, and how they are understood in the Langa case. This leads us to question the former's characterisation of actors, agency, motivating objects, and goals in relationship to the lived experience in Langa of business, violence, and development. We conclude in section 5 with directions for future research and practice, warning that the dominant binary characterisations of MSMEs and their roles risks obscuring the broader structural context, as well as suppressing the voices and aspirations of the most affected individuals and communities.

Contextualising Small Businesses in Violent Cities
Contexts of violence, fragility and conflict exist where 'political, social, security and economic risks correlate with organised violence' when 'states or institutions lack the capacity, accountability, or legitimacy to mediate relations between citizen groups and between citizens and the state'. (World Bank 2011: xv-xvi). This is increasingly true of urban areas, where violent and non-violent conflict have interlinked social, economic, and political causes (Winton 2004;Moser 2004;Moser and McIlwaine 2014;Kruijt 2007;Hoelscher & Nussio 2016). The challenges of urban violence in Latin America are well known (Balán 2002;Jütersonke et al. 2009;Pearce 2020) and are becoming increasingly important in African countries (Büscher 2018;Elfversson & Höglund 2019;Raleigh 2015;Urdal & Hoelscher 2012).
Businesses -particularly smaller and local private sector actors -are regularly victimised by urban violence. Extortion, robbery, kidnapping, and insecurity increase input relative to output costs, deter investment, constrict demand, reduce business confidence, affect where businesses are established and increase security costs and volatility of supply chains and operations (e.g., Goldberg et al. 2014;Rettberg 2008;Greenbaum & Tita 2004;Kimou 2015;Rozo 2018;Rosenthal & Ross 2010;Almeida & Montes 2020). Yet some business may benefit from violence and insecurity (Ajide et al. 2018), strategically seeking out violent markets (Bates & Robb 2008) or adopting a 'palliative approach' that has a limited interest in actual violence prevention given an inherent interest in a state of insecurity (Mathéy et al. 2014: 7).
In theorising MSMEs in violent cities, our review of the cross-disciplinary instrumental literature on business, violence, and fragility reveals a broad and somewhat disconnected series of scholarly endeavours. Concepts, terminologies, and representations of businesses in fragile or violent contexts differ across scholarly domains including political science, development studies, economics, law, sociology, and business studies. In synthesising, two dimensions -albeit imperfectly delineated -tend to reoccur. First, whether businesses intentions or impacts on peace and development outcomes are either productive or destructive; and second, the degree of agency businesses maintain such that they are active or passive actors. Drawing on this, one way to understand how businesses in violent contexts are conceptualised in the literature is using a framework of four stylised representations of firm types -predators, condoners, catalysers, and perserverers (Figure 1).
The four business 'types' emerging from this framework by and large align with examples in the literature (e.g., Goldberg et al. 2014) that show how SMEs engage through public and private mechanisms of collective action (catalysers), accept violence and increase security spending or adapt business strategies (condoners), directly engage in violence (predators), or curtail growth or even ultimately exit markets altogether (perserverers). However, as we discuss later in the article, these representations may not correspond with how businesses exist in real-world contexts.

Predators
Predators (active, destructive) are firms that are seen to willingly engage in or embrace the necessity of violence as part of their commercial operations. This may be either through an active and strategic use of violence, or a tacit recognition that close relationships with violent actors are required. Here the outcome is one which is negative -both economically and socially -as growth is stifled, resources are misallocated and violence undermines social and institutional stability and longer term development potential.
Much of this literature finds its entry point at the level of how businesses operate in wartime economies, engage in illicit markets, or act opportunistically during institutionally precarious conflict or post-conflict phases to negatively impact economies and societies. From early colonial and imperial conglomerates like United Fruit or the Dutch East India Company to the more recent entanglements of diamond or oil interests propping up parties in civil conflicts, predation and violence for private gain has long been linked with operations of Multinational Corporations (MNCs) in conflict or post conflict contexts (Ganson & Wennmann 2016). x Active acceptance DESTRUCTIVE This resonates in discussions of predation in war economies prolonging conflicts (Günduz et al. 2006b), and with warlords engaging in illicit business activities during conflict (Reno 2009). Predatory businesses are also commonly discussed in literatures around illicit economies and corruption (Naylor 2003). These businesses are seen to be 'peace negative' by exploiting and weakening institutional and rule of law deficits, using coercion and corruption for economic gain and political influence, often transforming to become semi-legitimate actors that mix licit, legal business interests with illicit and illegal economic activities (Cockayne 2010).
Closely linked with the notion of predation is the parasitic business (Torvik et al. 2006). Here, rather than being a business that engages in productive activities and contributes to the economy, 'parasites do not produce' and 'survive at the expense of productive activities ' (de Carvalho 2015). Parasitic businesses are seen as organised, violent criminal organisations or politically connected gangs that use or threaten violence for profit, and are linked with similar concepts including violent entrepreneurs (Volkov 2002) or destructive entrepreneurs (Baumol 1990). What these concepts tend to have in common is that in the absence of stable political and legal institutions, violence is able to be used by criminal and quasi-criminal businesses to engage in rent seeking and illicit earnings (e.g., Brück et al. 2011; and that this undermines productive economic activity. However, as Naylor (2009) points out 'there is little or nothing in the inherent logic of illegal markets to lead to the expectation of violence in their conduct' and that 'to the extent (that) violence does occur, the best place to look for its explanation is in the societal and political context' (Naylor 2009: 241).
Particularly for economists, violent entrepreneurship is deeply destructive for growth, squeezing productive businesses on both sides with the threat of violence and demanding use of their services for protection from violence (Mehlum et al. 2002). Here it is emphasised that 'in developing countries bandits not only control and protect small-scale informal enterprises such as street sellers and sweat shops' but that 'the surplus from large-scale modern firms are also dissipated by organized middlemen and professional plunderers' (Mehlum et al. 2000: 2, our emphasis). The result is an exacerbation of poverty, a stifling of productive entrepreneurship, economic stagnation, and the normalisation of violence for contract enforcement, where economies become mired in a 'Predators Club characterised by a vicious cycle of poverty and predation' (Mehlum et al. 2003: 275).

Condoners
Condoners (passive, destructive) may be seen as a grouping which accepts that violence is an inherent characteristic of their business operations and that they must tolerate its presence. While these businesses may not be violent themselves, and even bemoan the presence of violence, they recognise that they must engage with it in order to operate. This may include the payment of bribes and protection, selling to criminal organisations, or relying on informal relationships with illegal actors to ensure operational continuity.
This type of characterisation of the destructive firm is one that while they may or may not directly use or threaten violence, they condone its use or their operations rely on being embedded within a network of violent or illegal actors. This can often relate to legal and business perspectives around the nature of regulation and formality. The assumption is that unregulated, irregular, or informal businesses operate outside of oversight architectures and thus actively or passively condone the use of violence. Examples such as spaza convenience stores (Charman & Piper 2012) and shebeen liquor stores (Charman et al. 2014) in South Africa, and the mining sector in the Democratic Republic of the Congo (DRC) (Perks & Vlassenroot 2010) highlight how irregularity and informality in productive sectors is linked with persistent violence. Some of these conceptual linkages between the negative characteristics of firms in fragile contexts can be seen in the post-conflict private sector development literature, and are indicative of how the ' destructive firm' is seen in political economy and development circles.
According to Mac Sweeney, 'private-sector activity is often opportunistic and irregular, and lacks regulation or standardised systems of procedure', and though informality is seen to have some positive aspects, it 'will eventually pose major problems (for) government revenue collection and successful regulation ' (2008: 21). Furthermore, as in the case of the South African taxi industry (Dugard 2001), business actors step into regulatory voids to become 'informal agents of regulation, protection and extortion' that 'have been allowed to develop and expand virtually unchecked by the authorities', using 'their considerable firepower and weight to resist recent government attempts to re-regulate the taxi industry.' They 'are symptomatic of more generalised rising levels of organised crime in post-apartheid South Africa' (Dugard 2001: 3-4).

Catalysers
Catalysers (active, productive) can be regarded as the businesses that resist and fight back against violence, such as by refusing to pay bribes or sell to criminal networks. Beyond this, they may also organise and support individual or collective action to push back against violence in their communities. These virtuous firms are seen to have the potential to break cycles of violence and are the key to narratives of how businesses can support development, community building and peacemaking, often characterised as agents of positive change that the development community must support.
The catalyser narrative is one that sees businesses as an essential force for economic growth, employment and the assumed inherent good of formalised markets and commerce in fragile economies. In many ways this may be seen as an extension of the post-conflict private sector development wave of thinking that emerged in recent decades which has meaningfully shaped the discourse around the nature of businesses in fragile settings and their linkages with the dynamics of peace, conflict and development (e.g., Bagwitz et al. 2008;Guimond 2007;Mac Sweeney 2008;Naudé 2007). These representations broadly see firms as being able to use their capacities and relationships to either directly resist or oppose violence, or to create economic conditions that will reduce the risks of conflict emerging or being sustained. The productive firm view is also seen in accounts of firms that actively or passively challenge the proliferation of violence -aligning with how much of the Business for Peace, post-conflict development and DDR literatures see as the 'transformative' potential role of the private sector. Here, the broader economic returns to entrepreneurship are seen as ways to enhance intergroup cohesion or dialogue, and may therefore constitute platforms for conflict reduction or peacebuilding activities (Tobias & Bourdreaux 2011).
Other instances in the literature point to the roles that businesses play in resisting and opposing violence or supporting conflict resolution. Austin and Wennman (2017) outline the important role Kenyan business interests played in dampening electoral violence in 2007-2008 through mediation between conflict parties, petitioning political elites to respond in ways that deescalated tensions, and lobbying the media to soften inflammatory rhetoric. Similarly in Latin America, Hoelscher (2019) has highlighted the role of public utilities company Empresas Publicas de Medellín as a key financial and ideological supporter of successful public space and public security interventions in the city. Focusing on links between business, politics, and urban violence, Moncada has also noted how business elites align with political elites to support policies that have addressed urban violence in Colombia (2013).

Perserverers
Perserverers (passive, productive) can be viewed as those businesses that may be lesser targets for extortion or intimidation that are able to evade the attention of violent operators. They understand that violence is an inherent aspect of how or where they operate, but passively resist engaging with it, perhaps at the expense of their commercial prospects. They may, for instance, avoid forming relationships with violent actors, or limit size or growth in order to avoid becoming targets for extortion. Such passive, productive firms may understand the reality of engaging with violent actors within a violent systems, yet engage in forms of resistance or evasion in pursuit of non-violent goals, strategies Ramírez et al. (2015) describe as acquiescence and avoidance.
More recently, Moncada (2019) has examined how and why businesses challenge extortion from violent actors. He takes a political economy approach to show how firms resist criminal violence and rackets in violent contexts, identifying how the visibility of resistance can be either public or private, and the intent of resistance can be to overthrow a dominant power or negotiate the terms of domination. Here, a nuanced perspective on structural, subversive, everyday, and hidden resistance emerges, with the implication that strategies will depend largely on the political and economic resources available to firms. This highlights the continuum of active and passive roles that firms of different sizes and with different resource endowments can take to challenge and confront violence.

MSMEs, violence, and the grey zones of action and agency
On one hand, MSMEs are seen as highly precarious and vulnerable to violence and with limited political capital or access to finance to work through violence and grow their businesses. On the other hand, MSMEs are seen as critical partners in a dizzying array of development goals, including job creation, poverty, and inequality reduction, building trust and community cohesion, supporting state building and business formalisation, and acting as early warning systems for community conflict (Günduz et al. 2006a;Cordaid 2014). Yet it is difficult to tease out or generalise how and where these characterisations of MSMEs may apply, as the vast majority of mechanisms linking MSMEs, growth and peacebuilding remain unsubstantiated (Brück et al. 2015).
This diversity of roles is laid out in one of the few studies of SMEs in contexts of urban violence, where Ramírez et al. (2015) examine SMEs strategic approaches in Monterrey, Mexico. Rather than finding that businesses are necessarily 'good' or 'bad' actors, the authors highlight that SMEs differentially use approaches of manipulation, acquiescence and avoidance to contend with organised violence. Here, SMEs tend to remove themselves by relocation to less violent places, succumb to violence and accept the threats and coercion of violent actors, or become co-opted by violence and actively exploit it for commercial gain -but rarely do they actively resist violence. Similarly diverse strategies are noted in Joseph et al.'s (2020) study of MSMEs in Lebanon. They highlight that 'local business activity can simultaneously foster both peace and socially destructive outcomes' and that 'the importance of economic development was elevated for local businesses, suggesting that peace through mechanisms such as social development, the rule of law, and training, is only achieved if economic needs are alleviated through these measures' (Joseph et al. 2020: 1).
While the framework above outlines stylised portrayals of 'typical' business types, the literature also shows that these are not rigid categorisations -and many scholars in the field would likely suggest that they are imperfect. Yet while MSMEs are likely more fluid and diverse, these types of categorisations are still commonly deployed in the literature despite lacking real explanatory traction. We contend that this may be reflective of an underdeveloped theoretical understanding of MSMEs as embedded within systems of violence and the ways that this is understood in the context of communities beset by violence.
Rooted in the literature are normative and operational assumptions about the nature of violence, the structure and agency of 'firms', how they should or could behave instrumentally, and the resoluteness of their strategic relationships vis-à-vis violence and violent actors. In particular, the model of business types that we identify from the literature appears to imply an actor (the MSME) who has agency (to engage or not engage), and which undertakes (or fails to undertake) actions which are inherently productive or destructive with respect to the object of overt violence with the goal of either violence reduction and/or an economically-oriented notion of development. Yet as we outline in the following case study of Langa in Cape Town, South Africa, we conclude that this framework may inadequately describe the actors and their agency, objects, and goals as they are experienced and expressed by people in such places.

MSMEs in violent contexts: Experiences from Langa, Cape Town
Violence -even if embedded in broader sub-nation or national systems of conflict -presents within its own local context (Kalyvas 2006;Justino, Brück & Verwimp 2014). In order to understand the frameworks for understanding and frameworks for action that have currency at the intensely local level at which violence or development unfold, we therefore conducted a localised study of attitudes and experiences at the intersection of MSMEs and violence. We make no claim that the perspectives emerging from this one place are generalisable. Rather, we rigorously examine the understandings emerging from one well-delineated context so that they can be compared with the extent literature explored above, helping to uncover potentially uninterrogated assumptions and unanswered questions.
Langa, a suburb of Cape Town of roughly 52,000 inhabitants (Statistics South Africa 2011), was chosen because it is in many ways typical of similar suburbs across the municipality and South Africa, and of neighbourhoods of other cities around the world characterised by pronounced social, economic, and spatial segregation and inequality. The white government designed Langa in the 1920s on the then-outskirts of Cape Town as what both academic analysts (Coetzer 2016) and some residents describe as a ' concentration camp' in which control over black workers could be maintained. It was never intended to ' develop'; many of the streets are too narrow for two cars to pass, for example, because it was never expected that black residents would own cars and only police patrols needed to use them. Langa is a mix of brick and more informal structures, traversed by both paved roads and alleyways so narrow that two people cannot pass without touching. Langa remains intensely poor: as of the latest census, 72% of households had a monthly income under the official poverty line of roughly US $400; 40% of the population lived in a shack or other informal dwelling; under 40% of the population 15-64 years old was employed; only 40% of those aged 20 years and older had completed grade 12; and 65% of the population had no access to the internet from home, work, or elsewhere (Statistics South Africa 2011). With a murder rate as high as 69 per 100,000 inhabitants, Cape Town is one of the most violent cities in the world; and with a murder rate that may exceed 100 per 100,000 inhabitants, Langa is one of its most violent suburbs (SACN 2019). At the same time, Langa was historically, and today remains, a vibrant centre for political activism, social and economic self-help, and the performing and visual arts. It is a place of both immense pain and immense pride.
While the dominant dynamic of Langa for most of its history has been its forced segregation from, and deprivation in comparison to, the historically white areas of Cape Town, Langa also has a history of internal divisions that have contemporary repercussions. Langa was the product in 1927 of the forced relocation of black South African families from other parts of the City under the Natives (Urban Areas) Act No. 21 of 1923. The growing demand for labour, particularly in the Second World War, brought thousands more workers to Langa. These migrant men from rural areas lived in hostels, had no right to remain beyond their employment contracts in the City, and would be arrested even within Langa if they were found outside of the narrow corridor from the hostels to the train station. A divided suburb emerged, with separate shopping areas and social and cultural communities. With the Abolition of Influx Control Act (1986), some existing migrant workers brought their families to Langa, and many more arrived over the years, establishing sizeable informal settlements which in turn developed their own economies, social and civic structures and senses of 'us' and 'them'. Langa today is even officially divided into two Wards, corresponding roughly to ' old' Langa and to the 'new' informal settlements. Thus the dynamics of Apartheid divisions persist today.
Twenty interviews were conducted, supported by secondary research. The research followed an interpretivist paradigm (Alasuutari et al. 2008;Brand 2009) with the intent to not predefine dependent and independent variables. This was in recognition of the emergent nature of the inquiry around MSMEs and urban violence, and to ensure that the frameworks for understanding in the extent literature were not imposed on respondents. Selection was opportunistic, snowballing from multiple seeds (Morgan 2008) within the close contacts of one of the authors with activists in the community who then provided bridges to others. All the same, interviewees with diverse characteristics -across race, sex, age, class, profession, and sector (see Appendix I) -were consciously sought out, allowing for the possibility of divergent views and experiences within the primary data and a fuller complexity of human sense making (Kaplan & Maxwell 1994). At the same time, by selecting interviewees all of whom had experience of Langa, diverse perspectives could contribute to a more coherent systems perspective necessary to understand conflict and violence (Ganson 2019b). This methodology took into account both the need to locate members of a difficult to reach population (Spreen 1992) -in this case, people willing to talk openly about the specifics of conflict and violence in a context in which they lived and worked -and the recognition that the purpose of the study was not to establish a cohort that was representative per se, but rather to listen carefully to the narrative that emerged from the stories of one set of actors -in an in-depth and contextualised 'exploration of a central phenomenon' (Creswell 2005: 203) -that could be compared to extent theoretical frameworks.
The foundation of the inquiry was a primary, opened-ended inquiry: 'When you put three thoughts together -Langa, small businesses, and conflict and violence -what comes to mind for you?' While each interview may have used a number of other probes and follow-ups to maintain the conversation (see Appendix II), people were primarily invited to tell their stories. The goal was to allow participants maximum latitude to define for themselves the contours and boundaries of the inquiry. Their reflections are filtered through three lenses below: a narrative lens, using a series of vignettes that represent stories told that were typical of the broader interview pool; a thematic lens, drawing out from those vignettes and other primary data recurring observations and perspectives; and a comparative lens, probing points of intersection and divergence with the extent literature explored above.

Stories of SMEs, conflict and violence in Langa
The following vignettes highlight the reflections of different interviewees. They represent stylised stories that synthesise multiple conversations. First, individual interviews were completed; then, interview notes were harvested for common themes and story elements and the vignettes drafted; then these were discussed between a sub-set of interviewees for verification and refinement. The vignettes are presented here in part as entry points for reader into the rich world that is Langa, without the use, as a matter of conflict sensitive research (Svensson 2013), of quotations or unique stories that might identify informants. They are also concrete examples of the many rich expressions of 'violence', 'MSMEs', and the context of 'Langa' in which they unfold that bely any easy characterisation of the nature of, motivations for, or experiences of the interrelationships of small businesses with violence.

The street vendor
With money he has somehow saved or perhaps pulled together from family and friends, a man walks 10 km round trip to the city market across the railway tracks where he can buy not-quite-rotten fruit and vegetables at a discount. He must be up very early: the best deals are gone before 5:30, and in any case, his first customers may be on a taxi to work in the city centre before then. He knows that the police can steal his goods or money; or that on a given day price at which he can sell will not be better than the local corner stores. So he doesn't see himself as in the 'produce business'. He's in the business of identifying and seizing any opportunity that lets him turn his human and minimum financial capital into a small profit he can scrape by on -produce today, something else better if that comes up tomorrow. He's a friendly man, happy to chat between customers. But if someone insisted on selling at his regular spot? He'd have to kill him.

The development corridor
The municipal authorities from time to time advance plans for a road that would lead towards the airport. They argue that the stores that could be built along the improved byway would bring needed jobs and tax revenues. But the experience of local residents is that the profits from such ventures are extracted by the outsiders who have the capital for investment and the employees whom they bring in from their own networks. The project will thus undermine rather than build local business and employment. Additionally, the improved road would impinge on a green space where family members of all ages gather for picnics. If the young men who play sports there are forced away, where will they go and who will keep a watchful eye over them? The project is stalled in part because the municipality knows that the community will protect its interests with protests that history says can easily escalate into violence.
The taxi owner A retired teacher decides to supplement his inadequate pension. He combines savings with money borrowed from someone the authorities would call a loan shark to buy a minibus taxi. He pays a trivial fee to the city for his taxi license; and then between US $3000 and $5000 to join the 'voluntary' association that controls the apex industry in the township economy. He treads carefully; he knows he has entered the system as a privileged outsider, not having come up the ranks from fare collector to taxi rank manager to driver to owner. But he must also be firm. Taxis outside the association will always be trying to opportunistically steal his business by driving his route in front of him. This must be stopped, violently if necessary. This is in part because his lender, operating in a system outside of the constraints or protections of the rule of law for the small operator -borrower or lender alike -will feel compelled to make an example of him should he not repay his loan.

The immigrant
A refugee from a neighbouring country makes her way to Cape Town. Having a bit better education than most locals willing to work for such little pay, she finds a job in town. But spatial segregation of rich and poor means that she cannot afford to live anywhere nearby. She rents a shack in the backyard of a Langa home. She keeps to herself, both from lack of community connections, and from exhaustion. She discovers that she can buy chips cheaply in town and sell them for a bit more from home. Slowly she adds to her wares, and her shack becomes a store. Her landlord doesn't seem to mind; it makes her more dependable with the rent. But a robbery is explained to her as a warning to stop. When she doesn't, her shack is burned and she is chased away. The local consensus was that she took income away from a local resident who didn't have the same opportunities, and who without work would become a burden on the community.

The restauranteur
Having been discovered and trained by one of Cape Town's leading restauranteurs, he became a chef in his own right with a career in leading hotels around the globe. Now he has decided to return home to raise his own family where he grew up. He opens a restaurant that combines local dishes with a top-notch kitchen and professional front of house operation -even though guests sit at simple outdoor picnic tables. They are a mix of locals, tourists on township tours, and better off Capetonians who find it exotic to venture into a 'black' area for lunch or Sunday jazz. The owner takes pride in his high staff turnover; after all, he has taken in young people many would have called 'unemployable', trained them, and then placed them in fine restaurants across Cape Town. His business is safe. No local would attack him, embedded as he is in the community; and an outsider would do so at his peril.

The creche director
In her home, she used to care for children of neighbours who commuted for long days of work in the city's richer suburbs. She knew that the authorities considered this an 'illegal' unlicensed child care facility. But there was nothing she could do about the requirements for teacher certification or for ' adequate facilities'; and her brothers could be counted on to chase away any inspector arrogant enough to venture off the main road to cause trouble or seek a bribe. Then she by luck encountered a charitable organisation that provided her with training and set her up in a proper space, allowing her (after endless paperwork and delay) access to the small state subsidy as a registered early childhood development centre. As a business owner, she has confronted multiple break-ins and thefts to the centre. As a woman, she has been the victim of multiple sexual assaults, once by the police.

Emergent themes
From these stories, we can extract common themes that are echoed in the variety of voices of those who shared their experience and perspectives on Langa, small businesses, and violence. They emphasise business and violence as both being embedded, on the one hand, in a larger system of structural violence, and on the other, in a web of community relationships.

The dominance of structural violence
Interviewees report that the dominant, pervasive experience of violence in Langa is structural (Galtung 1969). The daily weariness of long commutes is a direct effect of Cape Town's unresolved spatial segregation; the regimented layout of Langa is a reminder of the need of elites to control black bodies and lives. Despair, drug abuse, hunger, domestic abuse, and a host of other social ills are seen as symptomatic of government decisions to deny decent education, withhold public amenities that are abundant in other areas of the city, and pursue low-employment, low-wage economic policies that benefit the few at the expense of the many. Furthermore, the police are a principal source of violence, engaging in theft, rape, and racketeering. Closely aligned with them in people's minds are the government agents who are largely seen to be sources of restraint and control, rather than of support and development.

People focused on survival
Interviewees explain that this context of structural violence forms the boundary conditions within which most people are simply struggling to survive. Becoming a 'business owner' -whether producing sandwiches to sell at the taxi rank, or seizing the opportunity to make a small profit with a corner store -is often not a conscious choice, but an imposed necessity. Any action that interferes with this vital mission of putting food on the table -irrespective of whether it is an overt act of aggression like theft, strangling government regulation, or an outsider attempting to enter a highly constrained economic space -is experienced as an act of violence. Responses to these actions are therefore not themselves inherently violent, but self-defence. Similarly, methods for conducting business that are considered violent can all the same be seen as subordinate responses to the failure of the broader system to provide protection or rule of law.
The primacy of networks of mutual support and obligation Interviewees underline that, within this context, people's primary -and sometimes only -source of strength and resilience are the social networks -family, church, neighbourhood, sports club, jazz club, or other -of which they are part. These webs of relationships provide people with emergency child or elder care, hunger relief, emotional support in times of trauma, a trusted source of short-term employees when there is temporary demand for more goods or services, and sometimes business finance. There are few if any boundaries between personal and professional identities. People do not support ' a business', but rather the person behind that business who is embedded in a social network of which they are both part. Similarly, if 'the business' is doing well, it is expected that the cash in its register be made available to support needs within the larger network.
The multiplicity of roles with respect to criminality and violence Interviews reflect that these factors, taken together, mean that there are few if any small businesses that act entirely apart from violence. Most people are victims: mobile phones are so frequently stolen, for example, that a woman may carry a back-up SIM chip in her bra. At the same time, they are in some way caught up in cycles of criminality and violence: that same woman may replace her phone with one bought at a price that makes clear that it must itself be stolen goods; or a small store owner might recognise the scourge of gangs but still give cool drinks to the young men congregated at his door, both to keep them from harassing customers, and to keep worse actors away. Violence within and between taxi associations may be condemned as excessive, while recognising that the taxi system as a whole provides some degree of order and a vital public service that to some extent reduce structural violence.

Challenges to the extant literature
These widely-reported dynamics explain a far more contextual and contingent framing of violence related to small businesses in Langa than the extant literature appears to recognise. People's perceptions of who the violent actor is, whether a response is violent, and whether such action is socially and morally permissible is filtered through at least three lenses: degree of social embeddedness, degree of local contribution, and relationship to structural violence (Figure 2).
The stories and thematic analysis suggest the characteristics of an enterprise that can be considered wholly peaceful. These are those, such as the restaurant described above, that are at the same time deeply imbedded within networks of mutual support and obligation; which are seen to be making a significant local contribution across economic (e.g., investment, local procurement, and job creation) and non-economic (e.g., pride in presenting local cuisine) dimensions; and which are neither dependent on, nor supportive of, state-sponsored violence, but rather stand in opposition to it. It represents a pinnacle of right relationships among people in Langa. By implication, any action which undermines or attacks such businesses is considered violent; and any action that is taken to protect them is not violent, but self-defence.
They also suggest the characteristics of an enterprise that will be considered wholly violent. Perhaps surprisingly to an outsider, so-called ' development' initiatives such as the roadway improvement will be experienced as such. The businesses they bring are seen as expressions of structural violence: externally driven agendas that undermine opportunities for, and right relationships among, the people of Langa across economic (store competition) and non-economic (common green spaces) dimensions. Opposition to such enterprises is therefore justified self-defence, and support of them by government, investors, or others is experienced as violent. These are in the same category as predatory police practices that operate outside of and undermine networks of mutual support and obligation; extract rather than contribute to the community; and perpetuate historical injustice and structural violence.
The vast majority of enterprises, however, will be perceived as mixed. There may be resentment against the more extractive forms of 'poverty porn' (Doward 2020), for example, as a form of structural violencericher and mostly white selfie-takers whose experiential entertainment visiting a township arises from the entrenched poverty of those whose home is Langa -but it tends to increase positive government attention and bring some resources to the area. Poverty tourism therefore operates largely unmolested. The actions of the taxi owner, the loan collector, or the neighbours against the immigrant store are recognised as violent, as non-violent dispute resolution is socially and culturally preferable; but assessments are balanced against the social embeddedness of the actors, the value they bring as part of the community network, and their alternatives in light of structural violence. These acts are thus more likely to be judged according to a complex algorithm balancing community interests, proportionality, and necessity.
This inductive framework for assessing the inter-relationships of SMEs with violence in urban settings derived from the Langa storytellers, different as it is, does not render the framework that emerges from the literature entirely wrong. It may to some extent be descriptively true for some enterprises in some places, with clear points of intersection, for example, between the 'predators' of the former framework and the 'wholly violent' of the latter. But to the extent that policy development or social change initiatives are premised on a particular understanding of the relationship of business to violence as set out in the extant literature -in relationship to concepts of actor, agency, object and goal as set out above -these community narratives suggest that the outside-in framework may not be broadly relied upon. Actor. Business activity in Langa is undertaken by people who describe themselves as being deeply imbedded in webs of relationships on which they are dependent for their social license to operate, commerce, and protection, and to whom they owe reciprocal obligations. It is socially, culturally, and economically infeasible -as well as possibly personally distressing or even repulsive -to operate outside of this communitarian ethic of mutual support and obligation. As importantly, the community network is the only institution that people have experienced as both effective and responsive to their needs. The unit of analysis for actors may therefore better be seen as the communal network than the business. Agency. The system of violence within Langa is widely understood as being a sub-system of the larger system of structural violence, limiting the freedom to operate for both an individual business owner and the networks within which she is embedded. Once one decides to enter into the business sphere of taxis, corner stores, or neighbourhood pubs, for example -the largest economic sectors by far in the township economy -one is part of that sphere including its inherent logics of violence; one cannot be part of the economic system and not part of the violence system. The agency problem may therefore better be seen not as one of choice, but of entrenched, negatively reinforcing systems.

Motivating object.
A physical attack on a street vendor may be described in Langa as wrongfully violent, self-defence (and therefore anti-violent), or justified violence. This turns on the position of the respective actors in relationship to social embeddedness (local or outsider?), local contribution of that actor (supporting or extracting?), and their relationship to structural violence (predatory police or neighbourhood watch?), as well as on the relationship of these factors to each other. The object in relationship to which people are motivated to act (or not) is therefore better seen not as violence per se, but rather the impact of any action or enterprise across these three dimensions -social embeddedness, local contribution, and reduction of structural violence -considered worthy of communal protection.
Goal. The stories of Langa are complex. They are on the one hand rooted in trauma: of historical injustice, the persistence of structural violence, and the sadness and even shame of communal institutions so degraded that they fail to intervene to stop domestic violence against children or the rape of women, both of which are acknowledged to be widespread. The stories are on the other hand rooted in strength and resilience: the jazz scene and political activism that flourished even under Apartheid, and the civic, religious, and social organisations that today promote art, community, and enterprise. The apex goal within this context may therefore be better seen not as 'violence reduction', but as right relationships between members of a community, from which peaceful development -including but not limited to reduction of overt violence -can naturally flow.

'Violence in' and 'Violence of' the System
The extant literature seems largely to support an inquiry into violence in the system: in particular, whether MSMEs are catalysers, condoners, predators, or perserverers, and how more might be moved into a catalysing role with respect to MSME agency and challenges to overt violence. The stories of Langa suggest the need for inquiry into the violence of the system: the ways the weight of history combined with current government policy and practice limits the agency of the poor, and forces people to adopt strategies of overt violence to oppose the structural violence that they experience as undermining the community cohesion and indigenous economic development on which a peaceful future for them is premised. The voices of Langa appear to be pronounced expressions of the particularly urban nature of violence they experience: the confluence of stark inequality, limited opportunity amidst heightened competition, diminished social cohesion and cultural institutions with rapid population growth and migrant mixing, and powerful public and private institutions that are expressions of factionalism, greed, and exclusion, rather than conduits for their eradication. This suggests that research, policy or interventions vis-à-vis overt violence need to simultaneously consider and take into account structural violence within a broader human security and peacebuilding framework.
Furthermore, the Langa stories expose a certain poverty in the concept of MSMEs itself. The vignettes do include one classic small enterprise -the restaurant -as well as microenterprises both informal (the street trader) and formal (the creche owner and her few employees). But the taxi driver, although formally a microenterprise, was fully embedded in the apex industry of the township economy within his particular taxi association, playing by rules not of his own making and with limited agency as a business owner. The same would be true for 'mom and pop' corner stores in Langa that were legally independent but all the same simply the last link in large, vertically integrated semi-licit business organisations on whom they depended for finance, equipment leasing, stock, and perhaps even protection. Other cases of small enterprises were recounted -for example, welders who had grown in reputation to where they could successfully seek business outside of Langa as well as within it -but whose mutual understanding with their work crews was that they were independent contractors, and who thus would be counted as 'micro' even as they moved towards 'medium' in actual size. In a related vein, the creche owner was nominally independent, but was wholly dependent on her participation in an NGO network for skills and expertise that allowed her to secure the public funding that sustained her business. Meanwhile, predatory police seemed at times indistinguishable from private gangs, and outside developers were seen as deeply entangled with powerful politicians and 'public' agendas. It is no wonder that there is little distinction in the stories told by Langa residents between the private sector, civil society, and government, or between micro, small, or medium firms, with regard to either peaceful or violent enterprises.
Mainstream business scholarship has for some time acknowledged that an enterprise is ' a set of relationships among groups that have a stake in the activities that make up the business' (Parmar et al. 2010: 5); that is to say, that the ' enterprise' is not coterminous with 'the company', as illustrated by the taxi association. There is also growing recognition of the need for, and emergence of, new models of social entrepreneurship, at least some of which intentionally blur lines between the private, the public, and the civic (Littlewood & Holt 2018), as illustrated by the creche. Implicit assumptions of a focus on MSMEs -for example, that the private sector character or size of a company is easily discernible, or particularly meaningful, or has a predictable impact on its agency within the broader system -may need to be questioned. In particular, the stories of Langa give credence to the perspective that, rather than larger enterprises having greater freedom to operate as catalysts for violence reduction, the way an enterprise becomes large in a place like Langa may be by further embedding itself in the system of structural (and sometimes overt) violence (Ganson 2019a).

Conclusions
MSMEs are seen as critical peace and development partners, and cities as central arenas for sustainable development. Moreover, MSMEs often face a unique set of opportunities and challenges in cities, having to negotiate interlinked systems and grey zones of formal and informal, licit and illicit politics, institutions, markets and social relations. In recognising the importance of both MSMEs and cities for peacebuilding and development agendas, this article has sought to address a simple question: Do the ways in which businesses in violent urban contexts are represented in scholarship correspond with how they are understood in reality? As our analysis of the literature and the case of Langa has shown, there are fundamental disconnects between how businesses in fragile contexts are conceptualised in academic and policy circles, and how they are seen as filtered through the lens of lived realities of at least one urban community beset by violence.
In identifying actionable guidance from these tensions, we see three central issues that must be reconciled to move scholarship forward and to improve policy and interventions that aim to support small businesses in delivering economic opportunities and fostering safer and more secure cities. Furthermore, since they each are calls for the contextualisation of efforts and localisation of agency in addressing violence and peaceful development, their validity is not dependent on the universality of the Langa experience -only on the disconnect of that experience from the extant scholarship.
First is the need to question the oversimplification or binary representations that pervade the literature. From an outside-in perspective, it is appealing, intuitive and perhaps comfortable to see SMSEs as il/legal, il/licit, in/formal, or un/regulated. Yet this framing ignores the intractability and embeddedness of violence as a constituent feature of both social life and business operations in violent communities. Violence is purposeful; it is 'utilised to achieve some goal and as part of a broader strategy' (Ganson & Wennmann 2016: 109;Keen 1997). In Langa, stories illustrate how it is purposefully applied by different actors at different times to support or to undermine the embeddedness of enterprises, to protect their social contribution or to permit extractive activities, and to defend against, compensate for, or enforce the dynamics of structural violence. From this 'inside out' perspective, violence is not a binary opposite of peace, but rather serves a range of purposes along continua of justified to unjustified, proportionate to excessive, destructive to peace promoting. Furthermore, the stories belie the implicit binary of the state as either present or absent. A strong thread through the Langa stories is the predatory state: criminal and violent police; government agencies that seek to suppress and control rather than to develop and support; and provincial and national policies that people experience as sacrificing the needs and aspiration of the community to feed the greed of those more powerful. As Langa citizens see themselves as engaged in an exercise of community survival and self-development, the very characterisation of their activities as illegal, illicit, informal, unregulated, or violent is seen as an expression of the predatory state and structural violence.
Second, a misunderstanding of such dynamics can lead to facile interventions that may at best do little to support peace-positive businesses or create safer communities, and at worst can unravel complex social systems and further exacerbate violence (Luiz, Ganson & Wennmann 2019;Miller et al. 2019). Outside-in goals to 'support non-violence', ' contain violence', 'incentivise entrepreneurship' or 'promote business for peace' can profoundly misconstrue the actors, agency, object and goals related to the use of violence, and thus the impact of ' anti-violence' or 'pro-business' interventions on community cohesion and peaceful development. Rather, violence and under-development ' can only be addressed effectively and sustainably if understood as symptoms of the broader structural context, particularly the reality of social relations and competing systems of power' (Hoffman & Lange 2016: 23). Here, the actions of individual firms cannot be delinked from the local context of violence in which they are embedded, and instead entrepreneurial activity and its impacts will be dependent on societal, institutional, and individual characteristics (Muhammad et al. 2016;Spedale & Watson 2014). Furthermore, as Austin and Wennmann (2017) note in their study of SMEs in Kenya, there is a need to move 'beyond a focus on "the company" in shaping the dynamics of conflict events -negatively by causing or exacerbating conflict; or positively by mitigating or resolving conflicttowards an understanding of companies as one of many actors and factors' in a conflict system and the political economy of MSMEs in contexts of violence (2017: 452; Ganson & Wennmann 2016). If interventions to assist MSME development in fragile settings fail to understand such complexities, they risk 'reinforcing the same structures that prevent small businesses from growing and fragile situations from stabilising'; and 'persistent attempts to transform the power configuration into a more equitable, transparent and fair business environment are likely to trigger strong opposition and potentially violent resistance' (Hoffman and Lange 2016: 52).
Third, and perhaps most fundamentally, is the imperative for the people with direct, lived experience of conflict, violence, small business enterprises, and development in their own communities to play primary roles in the shaping of research questions, data collection and analysis, and the development of the frameworks for understanding and the frameworks for action that emerge. There is growing recognition that, in order to be actionable and effective, research, policy, advocacy, and practice related to violence risk reduction and peaceful development must be aligned with the realities on the ground (Firchow 2018;Anderson, Brown & Jean 2012). Without making claims that we have adequately explored the context of Langa, or that Langa's experience is generalisable, the stories and insights from Langa underline that there is greater analytical leverage gained from adopting an inside-out approach that acknowledges and unpacks how MSMEs are a part of an ecosystem of different urban actors whose relationships and interactions create and produce spatially and temporally nuanced processes that outside-in approaches cannot possibly imagine or take into account. Furthermore, the voices of Langa make clear that the challenges of MSMEs in contexts of urban violence cannot be reduced to instrumental initiatives to contain violence, fix incentives, overcome technical constraints, or provide legal, financial or regulatory support. That is not to say that such measures may not play a role; but they must be subsidiary to conversations that honour local imperatives to recognise trauma, celebrate resilience, build community, and situate violence reduction and economic development within a broader and more humane context of community empowerment and healing.