Living on the margins: the social dynamics of economic marginalisation

Abstract

This paper introduces the key concepts and focus of a special issue of Development Southern Africa on the social dynamics of economic marginalisation. It emphasises the importance of the rise of ‘second economy’ discourse in South Africa, but warns against its implicit dualism: persistent poverty can flow not simply from disconnection but also from adverse incorporation. The contributions collected in this special issue highlight the diversity of ways the poor may be connected, disconnected or incorporated. The paper considers these ways and distinguishes between a number of different concepts of marginality. It also argues that the case studies presented in these papers show that policy-makers should beware of naïve or overly optimistic assumptions about the benefits of ‘integrating’ poor people into broader economic systems within which they have little leverage.

This special issue of Development Southern Africa focuses on the social dynamics of economic marginalisation in southern Africa. It presents selected papers from the March 2007 Living on the Margins Conference in Cape Town, an international event that explored the complex processes that shape life at the margins of the formal and mainstream economy and considered the implications for pro-poor and development policy in South Africa and elsewhere (see www.livingonthemargins.org).

This introduction surveys the underlying conceptual issues considered by these papers, provides an overview of the papers themselves, and highlights some of the issues that emerge from them. In addition, it reflects on the underlying shifts in official discourse and dominant policy debates that make a focus on economic marginalisation in southern Africa today topical and important – and possible in the first place.

One of the more significant shifts in discourse on growth, poverty and economic policy in South Africa since the transition to democracy has been the increasing official willingness since 2003 to recognise the limitations of the notion that the benefits of economic growth would automatically ‘trickle down’ to poor people. Since the adoption of the Growth, Employment and Redistribution (GEAR) framework in South Africa in 1996, policy debates have often been unhelpfully polarised. On the one hand, support for GEAR has been linked to an uncritical assumption that economic openness and integration into the global economy would create a rising tide that would lift all boats. On the other, this policy has been subjected to a critique that often (rather reductionistically) critiqued it simply as an expression of neoliberalism; that sometimes (perhaps nostalgically) harked back to the supposed virtues of the Reconstruction and Development Programme; and that tended to frame the debate simply in terms of a stark confrontation between a discourse focused on economic utility versus an insistence on the policy centrality of human rights. As these positions became entrenched, little real debate took place over the fundamental underlying assumptions about the relationships between growth, equity, poverty, inequality and social justice – and where these debates did happen, they often concentrated on scoring ideological points rather than on focusing on some of the room for manoeuvre that existed even within the constraints imposed by current macro-economic considerations.

Signs are that, in the past 2 or 3 years, space has been opening up for more creative approaches, and for these contentious issues to be somewhat more available for critical and constructive discussion in South Africa and abroad. In the field of development discourse generally, the increasing centrality of concepts such as ‘social exclusion’ and ‘adverse incorporation’ has to some extent challenged narrowly economistic arguments and focused attention on key aspects of social dynamics, power relations and social identity (Silver, 1994; De Haan, 1998a,b; Hickey & du Toit, 2007). Some would argue that the rise of ‘social capital’ discourse in development and poverty research also reflects at least an attempt to factor social process into economistic analysis (Baron et al., 2000; for a more sceptical assessment, see Fine, 2001). In econometric poverty studies, work on asset thresholds has called attention to the key links between chronic and structural poverty, and challenged neoclassical assumptions about pathways out of poverty (Carter & May, 2001; Carter & Barrett, 2005). Methodologically, the previously unquestioned centrality of economic and quantitative modes of understanding poverty is making space – albeit in a limited way – for mixed methods and qualitative approaches (Kanbur, 2002; du Toit, 2006; Kanbur & Schaffer, 2007). And in government discourse in South Africa, the increasing centrality of discourse on the ‘second economy’ has explicitly highlighted the reality of deep and persistent economic marginalisation and the segmented nature of the South African economy (Mbeki, 2003; Faull, 2005; du Toit & Neves, 2007). The notion of the ‘developmental state’ – which under the GEAR framework often seemed to play a mirage-like, ideological role – is, with the articulation of the Accelerated and Shared Growth Initiative for South Africa, somewhat more open to specification. Space exists for a more engaged and constructive discussion about the limits and possibilities of policies to address poverty even within the terms of the ‘post-Washington Consensus’, the potential benefits and risks of different growth paths with different redistributive effects (Nissanke & Thorbecke, 2005), and the nature and implications of present-day formations of capitalism in South Africa.

This increasing policy openness in South Africa is not just the result of new evidence. Despite the prevalence of nostrums about ‘evidence-based policy’, the reality is that policy-making is never in fact based entirely on data and evidence. This is not because politicians or policy-makers are short-sighted or irrational, but because policy debates, even when they appear to be value free and technical, are by their very nature ideologically structured. Policies are not simply normative decisions or agreements about the preferred nature of social change or the goals of collective or societal action; they are also embodied in assumptions about how problems are to be framed, and in the underlying and often unwritten rules about what is sayable and what is not, about which truths and realities should be invested with political consequence – and about what is, and what is not, up for debate and discussion.

It is in this context that we should see the significance of an increasingly dominant discourse around the ‘second economy’. Initially emanating from the South African presidency in late 2003 (Mbeki, 2003; Policy Co-ordination and Advisory Services, 2003; African National Congress, 2005), it swiftly moved into the mainstream of government and public discourse in the years immediately following, and is currently a centrepiece of government initiatives such as the Accelerated and Shared Growth Initiative for South Africa (see Faull, 2005). The notion of the ‘second economy’ as an economic realm existing alongside the ‘first economy’ but ‘structurally disconnected’ from it has of course been heavily criticised (Bond, 2007; du Toit & Neves, 2007; Frye, 2007; Masondo, 2007). Nevertheless, it should still be recognised as a significant discursive development: crucially, it allowed the limited recognition in official discourse of some problematic features of South African capitalism. In so doing it provided a way for official discourse and policy debate to ‘name’ (or rename) a set of socially and politically contentious issues and thereby include them in mainstream ‘policy talk’ – to group a set of rather disparate phenomena under a single heading that could be given more or less coherent meanings, ‘bundled’ together in political debate and targeted together in formal interventions.

This does not mean that the key ideological and political differences that animated the polarisation and divisions over issues such as the Reconstruction and Development Programme, the GEAR framework, the Basic Income Grant, service privatisation, social grants and land reform are much closer to resolution. What it does mean is that some of the underlying conceptual differences and issues have moved much more onto centre stage. The unquestioned self-evidence of neoliberal narratives about growth, employment and redistribution has been called into question. Key issues – deepening inequality, the limitations of ‘normal’ and mainstream economic growth, the possibility that the systematic marginalisation of millions of South Africans is not merely a holdover from apartheid policy but is being perpetuated by central and current features of the new social and economic order itself, the notion that economic growth as such and by itself will not alleviate and may sometimes even worsen poverty – are no longer so easily passed over, but are at the heart of mainstream policy and economic debate.

How this debate will go and what is possible within it is, however, far from settled. A key issue here is that the meanings and definitions of the key terms continue to be contested. Just what is meant by ‘structural poverty’, ‘social exclusion’ and the notion of the ‘second economy’ is itself the subject of contestation. Already it is clear that these notions can be appropriated from a variety of directions. On the political left, the concept of the ‘second economy’, if not rejected outright (for example, Frye 2007), can be used to critique the systematically exclusionary workings of the mainstream economy in a globalising world (Baumann, 2004), and as a synonym for the reserve army of unemployed labour dumped by twenty-first-century capitalism in Mike Davis's (2006) Planet of Slums. Within much of South Africa's business community it is seen uncritically as a synonym for ‘the informal sector’ and small, medium and micro enterprises. Within government discourse, the ‘second economy’ is sometimes seen as a problem sector to be ‘eliminated’ (Republic of South Africa, 2006) or integrated into the first – and sometimes as a potential powerhouse comprised, in De Soto-esque terms, of millions of frustrated potential or would-be entrepreneurs. There is a significant risk that discussions about the ‘second economy’ may become mired in vagueness and generalities, informed not by a detailed and clear understanding of the actual dynamics of inequality, exclusion and marginalisation, but by the prior assumptions and prejudices of whoever happens to be speaking.

With this in mind, this special issue of Development Southern Africa brings together a range of papers that highlight various aspects of the dynamics of economic marginalisation in sub-Saharan Africa. For obvious reasons, it is not an issue on ‘the second economy’: the ideologically contested nature of this concept and its broad application means there is no unproblematic way of defining what is and is not part of it. In many ways this is a residual and circular notion, loosely used to lump together whatever and whoever is not participating effectively or beneficially in the ‘mainstream economy’. For the same reason, it is not an issue on ‘the informal sector’ or ‘the informal economy’; although many of the sectors and enterprises considered here are informal, there are important parts of the informal sector that are not marginalised, while many who are ‘formally’ employed work in externalised and outsourced enterprises in conditions that are sometimes indistinguishable from those in the informal sector (Devey et al., 2006).

Rather than try to imagine this wide array of activities, sectors and persons as falling within a single ‘sector’ or ‘economy’, this special issue tries to consider the somewhat more abstract and flexible notion of economic marginalisation as such. What are its causes and consequences? What, in fact, does marginality consist of? How are ‘margins’ and ‘centre’ to be conceptualised? What are their (various) relationships with one another? And, crucially, is economic marginalisation purely an economic process, or do ‘social’ and ‘economic’ processes interact in complex ways?

In principle, such discussions should have a rich and varied theoretical literature to draw on in South Africa and beyond. Understanding the nature of inequality and economic marginalisation in southern Africa is, after all, not a new intellectual endeavour, and many contemporary discussions of the ‘second economy’ idea often traverse terrain that has been travelled before: indeed, as Deborah Potts observes, scholars may experience a sense of déjà vu as hoary debates and discussions on dualism are dusted off yet again. Contemporary discussions of the segmented nature of South African society and the limits of growth and integration often traverse – sometimes, it appears, unwittingly – terrain thoroughly explored by the wave of revisionist Marxist scholarship that overturned the certainties of optimistic liberal assessments of the nature of South African capitalism in the 1970s and 1980s (see, for example, Wolpe, 1972, 1980; Davies et al., 1976; Legassick & Wolpe, 1976; see also Legassick, 2007). These theoretical and conceptual resources are, however, seldom used: with some important exceptions (Crankshaw, 1997; Terreblanche, 2002; Seekings & Nattrass, 2005), class analysis has all but vanished from the mainstream of scholarly research about South African society. At the same time, as this author has argued at length elsewhere, the mainstream of English-language ‘Development Studies’ and ‘Development Research’ is similarly a-theoretical, narrowing poverty analysis down to the consideration of a small number of quantitative indicators, or to the increasingly apolitical, technicist and decontextualised application of a generic and power-blind ‘livelihoods model’ (du Toit, 2005).

In this context, and in the absence of a strong tradition of analysing social process and power relations in the fields of development and poverty studies, the most influential theoretical resource in the broader field of development studies for understanding the situations of those who seem to be economically marginalised or unable to share in the benefits of economic growth is provided by the discourse around ‘social exclusion’. (Variants of the ‘culture of poverty’ arguments persist in some quarters – for example, in the notion that the poor engage in a ‘Faustian bargain’ that involves trading long-term advancement for short-term security (Wood, 2003) – but these are not nearly as prevalent.) From its original roots in French social theory, where it referred mostly to situations where minority groups of one kind or another failed to integrate properly into the broader stream of civic life and citizenship (see Silver, 1994), ‘social exclusion’ has become a broad and all-purpose term denoting situations in which a broad range of factors – from physical and mental disability or chronic illness to gender, age, identity, caste, race, unequal power relations, discrimination, political bias, the nature of the state and the nature of regional and spatial integration – prevent individuals or groups from participating fully in social, political and economic life (De Haan, 1998a,b; Kabeer, 2002; Beall & Piron, 2005; Hickey & du Toit, 2007).

In South Africa, the reception of ‘social exclusion’ discourse has been complex. As some earlier commentators have pointed out, sub-Saharan Africa, characterised by the relative absence of strong traditions of civic identity and citizenship, offers little purchase for the language of ‘social exclusion’, which usually focuses on minorities defined by their relatively exceptional exclusion from the terrain of mainstream citizenship (Gore, 1994). The term has sometimes been used to show how identity can be used to exclude people from certain kinds of right and entitlement (see, for example, Beall, 2002). For all that, explicit references to the notion have in South Africa until recently been confined to rather specialised policy and technical circles.

On the other hand, there are strong thematic resonances between the discourse of social exclusion and broader streams of political and economic thought in post-apartheid South Africa. Indeed, one of the remarkable features of political and social discourse in South Africa across a fairly wide political spectrum is the centrality of notions of inclusion or exclusion in the conceptualisation of poverty and inequality. Thus ‘mainstream’ political and economic discourse is predicated on the notion that growth requires openness, liberalisation and integration into a ‘global’ economic order, while poverty is explained as the result of the failure to participate or be included in this economy. This is the assumption obviously at work in the discourse around the ‘second economy’; it is also a central trope in the narratives about Africa and its place in the world economic order that inform initiatives like the New Partnership for African Development (2001) . Interestingly enough, however, similar assumptions also often seem to inform the discourses of those who explicitly see themselves as critical of this policy mainstream. Thus many of the ‘new’ social movements that have arisen in South Africa since 1996 are predicated very much on a discourse of exclusion;1 the difference being that what is sought is not so much inclusion in the capitalist economic order as inclusion and integration in a polity and social order defined by human rights. In essence, both these discourses are dependent on a powerful but often unexpressed set of underlying assumptions about the implications and benefits of modernity and modernisation: while ‘liberals’ and proponents of economic growth see economic participation as the privileged vehicle of modernity, their social movement and left-wing critics privilege the forms of modernity inaugurated by citizen organisation and the protection of rights by the state.

Again, this author has elsewhere dealt at length with the strengths and weaknesses of ‘social exclusion’ discourse in accounting for persistent poverty in a southern African context (du Toit, 2004a). For the purposes of this special issue, the following salient points should be noted. On the one hand, ‘social exclusion’ discourse can arguably play a meaningful role in re-introducing a focus on the complexities of social process and social antagonism into the often rather power-blind and socially thin accounts of persistent poverty provided by economistic explanations. In particular, it is valuable for the way it draws attention to those occasions when people are not able to participate to their own benefit in the dominant streams of local social and economic life. On the other hand, the notion often poses more questions than it answers. It is often used in deeply divergent ways, and can be used to name a wide range of often dissimilar phenomena. Perhaps most importantly, naïve use of the term often seems to presuppose a simple opposition between ‘inclusion’, simplistically assumed to be good, and ‘exclusion’, which is equally unreflectively seen as bad. This reductionism ignores the situation of many of those whose marginalisation consists in the fact that they are included on disadvantageous terms. The notion that it is necessary to focus not just on exclusion as such but also on the terms of inclusion has led to a growing body of work that supplements social exclusion discourse with the notion of adverse incorporation (Beall, 2002; Murray, 2002; Bracking, 2003; Hickey & du Toit, 2007). While these are relatively novel ideas in the field of livelihoods research, they have of course been explored in detail elsewhere in the study of commodity systems and global value chains (see, for example, Gereffi, 1994; Friedland, 2001; Gibbon & Ponte, 2005).

Elsewhere, this author and David Neves have explored the implications of such a perspective for the notion of the second economy (du Toit & Neves, 2007). Considering the dynamics of jobless de-agrarianisation in the context of the ‘deep rural’ Eastern Cape, we have suggested that poverty is not a simple matter of the region's ‘disconnection’ or ‘lack of integration’ with the mainstream economy – indeed, the local economy is thoroughly penetrated by the lead firms of South Africa's banking, telecommunications and retail sectors. Rather, the nature of these connections acted to crowd out (and thus ‘exclude’) small retail enterprises and marginalise small-scale farmers. The same complex overlay of exclusion and disadvantageous forms of inclusion, we argued, were at work in perpetuating apartheid's spatial legacy in Cape Town's African suburbs. Neither the causes of poverty nor the limited forms of economic activity that were in fact possible could be well understood by looking at them through the lens of the ‘second economy’ metaphor. Rather, these cases illustrated how both margins and centre are everywhere present in the South African society, fractally interpenetrating and overlaid on one another. Although the notion of the ‘second economy’ was useful in so far as it placed on the agenda the importance of understanding the nature and terms of people's adverse structural location within a South African society conceived as a larger, segmented, and spatially articulated whole, the adverse aspects of this location were related not only to disconnection but also to the problematic nature of the connections that did exist between margin and centre, formal and informal, and the very direct linkages whereby costs, resources, shocks, opportunities and risks were transmitted from centre to margins.

Thus understood, the notion of adverse incorporation (and of economic marginalisation) is useful in much the same way as the notion of social exclusion: it functions as a kind of broad heuristic device, directing attention to the social processes and power relations that dictate the terms on which people are able to participate in social and economic life. As in the case of social exclusion, adverse incorporation and marginalisation can be used as portmanteau terms that cover a wide range of empirically very different situations. Crucially, they require answers to many more key questions. When is incorporation adverse, and when is it not? What are the underlying dynamics and processes that drive it? And, again crucially, is adverse incorporation to be understood in exceptionalist terms – as an accidental imperfection, resulting, for example, from market distortions, in a broader economic and social system that is in itself essentially neutral – or are adverse forms of incorporation systematically and unexceptionally created as a matter of course by the normal functionings and tendencies of a market economy?

Such questions cannot be answered on the basis of first principles or sweeping generalisations. They require careful consideration of a wide range of different contexts, sectors, activities and processes within which dynamics of economic marginalisation are at play, or wherein some kind of sustained and systematic patterned relationship between a ‘centre’ and a ‘margin’ can be proposed. If vacuity is to be avoided, the proponents of adverse incorporation as an explanatory category cannot simply assume it to be present in any context where persistent poverty and inclusion co-exist. Rather, the precise terms of incorporation have to be taken into consideration in every context. Similarly, the notion of economic marginalisation would be meaningless if it was simply expected to be a self-explanatory category: rather, attention has to be paid to both the ‘centre’ and the ‘margins’ – and the relationships between them.

This is the challenge to which the present special issue in some measure responds. Drawing on a range of contexts, it seeks to explore some of the different ways the relationship between economic centre and margins can be articulated.

It starts with an overview by Deborah Potts of the history of dualist conceptions of the relationship between the formal and the informal sectors in sub-Saharan Africa. As she emphasises, many of the issues raised by the discourse of the ‘second economy’ in South Africa have been exhaustively discussed elsewhere. After reminding readers of some of the key insights from earlier debates, she describes the changing fortunes of the informal sector in development and policy discourse: how it was initially seen as a problematic, laggard and tradition-bound sector; how the advent of neoliberalism saw it reconceptualised in more positive terms as a major employer and a repository of entrepreneurialism; and how some African states in recent years have again started seeing it as inherently problematic. Her account is also useful in that it charts the uneven impact of neoliberalism and ‘free market’ policies on the informal sector, and how this often worked to the advantage of big capital (shopping malls, industrial beer brewers) at the expense of small entrepreneurs. Her account also shows the importance of a comparative and regional perspective, highlighting, for example, the differences between the Zimbabwean Government's draconian actions against informal sector traders in that country and the rather more ambivalent attitudes reflected in South African discourse on the ‘second economy’.

Patrick McAllister's paper deconstructs the simple bipolarity between ‘centre’ and ‘margins’ and between the ‘first’ and the ‘second’ economy in livelihood formations in rural South Africa. He questions the simplistic notion that marginalisation equates to disconnection with the metropolitan economic mainstream. His study of Shixini shows how effective accumulation and livelihood strategies depend not in the first place on ‘effective integration’ with the metropolitan economy but on social relations and practices that are in part sustainable because, in contrast to the situation in other parts of the Eastern Cape, they have not been disrupted and reorganised by connections with the market economy. Being ‘marginal’, McAllister implies, is not necessarily a bad thing; or rather, we should be careful not to make assumptions about what is at the margins and what is at the centre. Indeed, understanding rural livelihoods in Shixini requires that we study livelihoods with a lens that puts Shixini at the centre. From such a perspective, we see that those who are ‘marginalised’ in Shixini are not disadvantaged because they are stuck in a ‘second economy’, disconnected from the ‘first world economy’; rather, they are marginalised from Shixini – marginalised because they are unable to participate in local practices of reciprocal exchange and instead have to depend on distant sources of income (social grants, remittances) instead of the activities upon which the local sense of moral community is founded.

These analyses, which relativise and deconstruct simplistic conceptualisations of marginality and disconnection, are followed by papers that interrogate and problematise some of the key dynamics of integration. Barbara Tapela analyses the ‘revitalisation’ of irrigation schemes predicated on the entry of small black farmers into commercial farming in Limpopo Province. She shows how water reform depended on the integration of informal, subsistence forms of agricultural production into commercialised agriculture. Her paper strikingly depicts the political and policy overdetermination of these interventions. These were not simply geared to helping small farmers by getting irrigation schemes to work; rather, they were intended to ensure a ‘holistic approach’ to ‘socially uplifting’ smallholders, ‘capacitating’ them to farm ‘profitably’ and ‘sustainably’. The term ‘revitalisation’ came to mean, at one and the same time, the technical reconstruction of ‘dilapidated infrastructure’ and the market integration of supposedly ‘disconnected’, ‘backward’ and ‘non-profitable’ small-scale farming into market-oriented commodity systems. Water reform, it seems, was here driven by a teleological and modernising policy discourse that sought to reconcile – and appeared to assume congruence between – the divergent imperatives of land reform, agricultural liberalisation and modernisation, commercial integration, globalisation, livelihood generation and ‘sustainability’. Her analysis reveals the unintended consequences of initiatives that on paper seem to square all these different policy circles coming unstuck in the messiness of reality. Poorly thought-through interventions that did not pay attention to interests and social relationships on the ground led to intractable conflicts. Small farmers without the financial wherewithal to compete effectively in commercial cotton production ended up indebted and in danger of losing their land; some have chosen to focus on less profitable but less risky local markets rather than the globalised commodity chains to which planners imagined they would be linked. The notion that water reform had to pay its way ultimately meant that vulnerable and monetarily poor farmers were subject to greater degrees of risk, as well as stringent phytosanitary requirements and other regulatory pressures with which they were poorly positioned to comply.

A slightly different picture is painted by Eleanor Fisher, who considers the dynamics of integration and formalisation in a very different context: that of artisanal mining in Tanzania. Again, she shows how certain kinds of marginality can be advantageous, how certain groups of people are able to ‘use’ their marginal status to exploit the ambiguous realm between legality and illegality, and how the relative freedom created by poor governance processes means advantages for some. She also brings to our attention the significant impact of power relationships and socio-economic disparities in contexts of relative informality and greater integration. She shows how mining was governed, not simply by the formal system of ownership and licensing, nor by a ‘disconnection’ with such systems, but by a complicated mixture of formal and informal. Integration and formalisation have also had a differential impact and intersected in complex ways with the existing patterns of advantage and disadvantage: while for some exclusion was reduced, for others disadvantage and marginality were intensified. Her article suggests that neither marginality nor ‘integration’ should be seen either uncritically as a good in itself, or as negative; rather, attention has to be paid to the way integration affects existing dynamics and power relationships.

The complex dynamics of how regulation shapes both the inclusion and exclusion of self-employed people is next considered by Richard Tambulasi and Happy Kayuni, in their discussion of the Malawian Government ban on minibus ‘callboys’, and by Caroline Skinner in her analysis of the ‘struggle for the street’ among informal traders, formal businesses and municipal authorities in Durban. In contrast to the earlier papers, which are concerned with the implications of integrating marginalised and informal activities into the economic networks of mainstream market economies, these papers take up Potts's explorations of the ways states and governments can engage with ‘the informal’. Tambulasi and Kayuni recount how the informal and unrecognised status of minibus ‘calling’, initially allowed to flourish in a ‘grey zone’ of tolerated illegality, rendered it vulnerable to government clampdown. Skinner explains how a key aspect of the relatively more progressive policies towards street traders pioneered in Durban during the period 2001 to 2005 was the way the city accepted the reality and importance of the informal sector, not as something that had to be transformed into something else, or ‘integrated’ into the mainstream economy, but as complex and important in its own right. Her account of inconsistent policy implementation and the crackdowns to which street traders have been subjected in recent years again shows how continued policy ambiguity can make informal traders vulnerable. It also highlights the persistence of official antagonism to the apparently ‘unregulated’, ‘unmodern’ aspect of informal trading, and the way discourses of modernisation and the ‘world class city’ can backfire on the urban poor.

The papers presented in this special issue cannot hope to be an exhaustive treatment of all the different modalities of economic marginalisation, or all the different possible ways those ‘at the economic margins’ are situated in relation to the mainstream. What they do achieve is to call into question any simplistic understanding of the centre-margin dichotomy, and in particular they problematise the optimism of neoliberal or technicist approaches to development for which progress is just a matter of linking poor people to the opportunities perceived to be flowing from a market. At the same time – and as crucially – they also problematise the sweeping notion of a world polarised into a globalising, capitalist ‘centre’ on the one side and millions of excluded ‘wasted lives’ on the other (Bauman, 2004). As a counterpoint to such views, these papers highlight the diverse ways people and their livelihood activities are inserted into the broader economic and social orders that shape their lives.

This raises an important conceptual point: rather like the notion of ‘exclusion’ to which it is to some extent related, the notion of ‘marginalisation’ lends itself easily to circular or tautological uses, in which ‘marginality’ becomes an unexamined, catch-all explanation for any particular situation of disadvantage. If this is to be avoided, policy-makers and researchers will have to pay careful attention to the details and specificities of the social relations, institutions and practices that mediate people's participation in social and economic life, and consider the precise extent to which these are enabling or disempowering, advantageous or adverse.

Analytically, this means we must abandon the dualism whereby economic activities are seen as either ‘(well) integrated’ into mainstream economic activities, or not. Any particular economic activity is always connected in a number of ways to a range of broader social, institutional and economic formations. Some of these connections are to the commodity chains, markets and institutions of the formal economy, and some are not. Homesteaders in Shixini are not accurately seen through a lens that focuses simply on their disconnection from the broader market economy; rather, what matters is how far they can participate advantageously in the local moral community. Furthermore, informal economic activities can never be reified as such or understood in a vacuum: as Potts's analysis makes clear, the formal and informal sectors in any economy are closely linked – indeed, formal and informal sector activities often subsidise, complement or are parasitic upon one another.

It should be clear that it is not possible to say a priori that integration into the formal institutions and systems of the mainstream economy is as such either advantageous or adverse. Both formality and informality can offer advantages and disadvantages: being connected to the economic mainstream could lead to poverty, and existing at its margins could bring prosperity. Which way it goes depends on the nature of the economic activity and the social situatedness of the individual or group under consideration.

This is not of course to suggest that the answer to any question about the nature and implications of marginality is simply ‘it depends’. An emphasis on particularity and local complexity does not have to plunge us into mere empiricist exceptionalism, nor does it mean that nothing systematic or general can be said about the implications and consequences of either ‘marginality’ or ‘integration’. Papers such as those presented here make it clear that there are broad lessons to be learned from specific instances.

Firstly, they highlight the range of possible kinds of ‘marginality’, their good or bad implications for the people concerned, the complex roles played by formality, informality and formalisation, and finally the need for conceptual clarification. Powerful and compelling though the notions of the ‘second economy’ and ‘economic marginalisation’ may be, a number of different issues and meanings are at play here. It appears that the notion of ‘marginality’ invokes two related but distinct meanings.

One is the degree or extent of integration and connection. Someone or something is marginal in relation to a particular field or formation if it is not strongly connected to, tightly integrated with, or centrally positioned in it. In this sense, marginality is not necessarily disadvantageous. For people who would otherwise have little bargaining power or clout, some degree of externality to the system may decrease risk or buy them room to manoeuvre. For some, like the artisanal miners in Fisher's article, existence at the edge of the economic and legal system may create opportunities and benefits that would not otherwise exist.

The other, however, is a person's lack of power or leverage in a given context. These two kinds of marginality do not necessarily coincide. Fisher explains how tighter integration into formal titling and licensing systems brings advantage for some miners but decreases the leverage and opportunities for others who survived as artisanal miners before formalisation or who have been attracted to the activity in recent years. And the small farmers Tapela describes are indeed caught between the two kinds of marginality – either they can try to participate in commercial agricultural activity, becoming tightly integrated into an economic system in which the risks are high and their leverage is scant; or they can choose to disengage from this system, becoming more ‘marginal’ but also winning some freedom and reducing their vulnerability.

As these examples show, neither of these meanings of marginality should be confused with the notion of informality, which is a third axis of meaning in the cases considered in these papers. The situations that the discourse on the ‘second economy’ seeks to name can therefore be mapped as points on a three-dimensional Cartesian space (see Figure 1): at the one extreme, denoted ‘a’ in the diagram – and characterised by formality, high degrees of integration and considerable leverage and resources – would be the firms and enterprises characterised in ‘second economy’ discourse as belonging to the ‘modern first world economy’. The impoverished villagers in McAllister's account would fall at the opposite extreme, denoted position ‘C’ in the diagram, disconnected from the formal economy and lacking leverage and resources in their local context, and would thus be classical examples of those whose poverty is linked to their exclusion. But many other permutations are possible. Those described in the papers in this special issue who derive some advantage from informality and low levels of integration into local or global value chains would lie at point ‘D’ in the diagram, while Tapela's farmers, incorporated into commercial agriculture on disadvantageous terms, would be at position ‘b’.

Figure 1: Mapping formality/informality, leverage and integration in three-dimensional space

These distinctions may help analysts to untangle many issues in the discourse on the ‘second economy,’ which so often conflates economic powerlessness, lack of connection and informality. In seeking to identify sources of advantage and disadvantage, researchers should not only ask whether particular people are connected to the ‘first world economy’ or not, and should not assume that all those in the ‘informal economy’ are disconnected (they are not). They should also take into account the nature of the leverage and the kinds of socio-economic power available to people in different contexts (‘integrated’ or ‘disconnected’, formalised or informal).

Secondly, at the theoretical level, the papers deal with some key recurring themes relating to the processes and dynamics that shape the kinds of power and leverage available at the margins (in both senses) of the formal economy. In particular, they spotlight the central importance of the social dynamics of economic marginalisation. They show powerfully how both advantage and disadvantage are not simply an outcome of the way disembodied economic actors are placed within the institutions and mechanisms of perfect or distorted markets; rather, markets, economic institutions and forms of exchange are socially constituted; and shaped by identity, power, ideology and meaning as much as they are by the economic realities of supply and demand. Understanding the implications of disconnection or integration, and gauging the extent to which these conditions are empowering or the reverse, therefore requires us to connect social and economic analysis.

This is a complex and exacting task. Elsewhere, this author and others have developed a simple heuristic framework to be used in research that aims to understand the terms of incorporation into broader economic circuits and processes (du Toit, 2004b; Bolwig et al., 2008). We have proposed that analysis should, as it were, proceed along two axes. On the one hand, it should pay attention to what might be called the ‘horizontal’ axis of local sociality: the role played by local culture, relations, processes and institutions in mediating the impact of broader economic processes. At the same time, as Colin Murray has emphasised, livelihoods should be contextualised in a broader political economy (Murray, 2002): alongside the dimension of local sociality, analysis should look also at the ‘vertical’ links emphasised by filière or global value chain analysis: the upstream and downstream links within particular value chains by which local actors are often connected to distant processes and institutions (Gereffi, 1994; Gibbon & Ponte, 2005). To understand the vulnerability of, say, seasonal fruit farm workers, we need to understand the re-articulation of paternalist labour regimes and the implications of increasing buyer-drivenness in the deciduous fruit value chain. The extent to which anyone benefits from participating in particular forms of exchange or production is given not only by their economic efficiency but also by who they are, and by the local meanings, histories and antagonisms that inform and shape these practices. McAllister's and Fisher's papers describe cases in which local economic relations that, from the state's point of view are ‘informal’, sustain and inform vital livelihood activities. At the same time, they warn against the simple ‘valorisation’ of social capital: some are better positioned than others in the social networks that sustain and are sustained by these forms of exchange. For those disadvantaged (for example) by age and gender, and for those who are not granted political or social authority by local tradition, the terms of exchange can often be punishingly unequal.

Thirdly, these papers reveal some clear empirical patterns that are of relevance to policy debates. Although simple teleological, dualistic or a priori assumptions about the benefits or disadvantages of integration or marginality need to be avoided, the case studies do provide some evidence about who typically stands to benefit and who does not, and about how particular forms of disadvantage come about. In particular, they repeatedly draw attention to the uneven impacts of integration, and the likelihood that those who are socially disempowered or who lack resources to begin with will be incorporated into the value chains and systems of the mainstream economy on terms that are unfavourable to them. ‘Adverse incorporation’, although not the universal consequence of participation in the formal economic system, is not for that reason to be understood as an exception to or aberration of that system.

Little comfort is thus given by the papers in this special issue to the notion that persistent poverty or marginality is ‘residual’ in nature, merely the consequence of distortions or imperfections in markets that, were they to develop fully or be given free rein, would allocate resources efficiently. The immiseration of the female pensioners whose fate Tapela describes, the struggles of the marginalised artisanal miners in Fisher's account, and the marginalisation of the Zambian street traders considered by Potts are not just unfortunate historical accidents – they are the systematic outcomes of the ‘normal’ workings of the commodity systems, tenurial formations and planning priorities that were pivotal to the different development policies at the centre of each story. Above all, these stories illustrate what goes wrong when policy-makers or development practitioners seek to integrate people into the broader, global economic order without paying attention to the real dynamics of power involved in that integration. After all, the forms of ‘marginality’ these papers describe are not historical accidents but the outcomes of complex historical and political processes. Those who make a living as informal street traders in Durban; those who rely on the gleanings of the Eastern Cape's smallholder cattle sector; those who subsist in the interface between the formal and the informal in the Tanzanian mining sector; the rural women pensioners in Limpopo – all these are already multiply disempowered, and are often relegated to what Ferguson describes as ‘survivalist improvisation’ because of the lack, or closure, of other opportunities (Ferguson, 2007). If economic programmes are to allow such people to become economically active in ways that are to their advantage, the complex and unequal power relationships that exist both ‘horizontally’ and ‘vertically’ have to be taken into account.

Incorporating these insights into policy debates and pro-poor programmes is a challenging task, not least because these are not merely technical matters. The key issue to be raised here, after all, is the nature and implications of the form and organisation of twenty-first-century globalised capitalism. Within the mainstream of the South African policy debate, framing the question in these terms has been in some sense verboten; capitalism itself has since 1994 no longer been a legitimate object of critical concern. At the same time, critique of this mainstream has often been content to reduce ‘capitalism’ to a swear word, and ‘neoliberalism’ to a catch-all term of dismissal that is assumed to explain all the sins of macro-economic and development policy. Clearly important questions about the immiserising impact – and the environmental unsustainability – of present-day capitalist economic formations remain. But significant space also remains for useful and important policy debates about different growth paths and policy scenarios even within the bounds of what is macro-economically possible. Scope still exists for the notion of a ‘developmental’ state in South Africa to be given real and consequential content. But such discussions are only possible if current debate goes beyond the teleological and dualist notion that poverty is due simply to people's disconnection from a mainstream economy that is not itself subjected to critical investigation, and begins to pay attention to the real-world institutional arrangements, social processes and economic dynamics of a single, differentiated, globally integrated and unequal economy.

Notes

¹I thank my colleague Jan Mogaladi for this insight.