5 | THE SOCIAL EXCLUSION APPROACH

The other major concept in poverty studies that has stood the test of time and of fleeting fads of self-promotion is that of social exclusion. The concept emerged in the 1970s and has become a central reference in the lexicon of poverty studies since the 1990s, to the extent that poverty and social exclusion are often used inseparably. It is one of the three approaches dealt with in this book because of its prevailing usage as a poverty concept in official policy documents, especially in the UK and Europe, and also in scholarship the world over. It is also the antithesis of social inclusion, which has become a cornerstone theme for the World Bank as well as for current global development agendas. In this sense, common definitions of inclusion within the discourse of inclusive growth or development are essentially mirror images of previous attempts to define exclusion in the 1990s.

Both concepts, however, are politically contested and suffer from vagueness, even while they garner wide appeal. Early discussions around inclusive growth or development in the late 2000s revolved around the tension between absolute poverty reduction (roughly represented by the World Bank position) and relative poverty or inequality reduction (roughly represented by the UN position), thereby reproducing the frame of debate in the idea of pro-poor growth that preceded it in the early 2000s. However, within this frame, inclusion is definitely juxtaposed with the idea of exclusion, by both the World Bank and the UN. This is clearly evident in what appears to be the most recent effort by the World Bank to give substance to the idea of social inclusion (WB 2013), as linked on their website at the time of writing.¹

Debates around social exclusion in the 1990s nonetheless offered more nuance. The exciting potential presented by the concept was the chance to bring in subtler sociological analyses into the existing field of poverty studies, such as notions of social integration, ordering, segregation, stratification and subordination. It is also commonly said that the attractions of the concept of social exclusion are that it brings attention to relational dimensions of poverty, highlights the types of processes that cause entitlement or capability failures (as discussed in the previous chapter), and brings to the fore issues of discrimination or disadvantage in the discussion of poverty. For these reasons, it is generally valued as a concept, even though it is never very clear how, exactly, it should be differentiated from poverty. In the policy literature, such as in documents by the European Commission that have officially adopted the concept, it is generally defined as a type of poverty.

This leads to the perennial question of whether the concept of social exclusion adds any value if it is essentially synonymous with poverty. Why not just talk about poverty? Amartya Sen (2000) argues that the concept is essentially redundant, perhaps semantically useful for the reasons mentioned above, but already implicit within various existing approaches to studying poverty, especially the capability approach that already deals with issues of relationality. Social exclusion can therefore be seen as an adjunct to this and other poverty approaches, describing various contextualised social causes and/or social consequences of poverty, albeit with an extra emphasis on coercion and discrimination than is usually made in the more liberal strands of the poverty studies literature. After a period of debate about the usefulness of the concept in the 1990s and early 2000s, many seem to have settled with this compromise.

Others, however, have raised more polemic critiques. The identification of exclusion as a principle characteristic of poverty encourages policy approaches that seek inclusion as the solution. Mistrust with the concept has been due to the fact that this is usually interpreted more specifically as market inclusion, thereby harmonising with neoliberal obsessions of extending and deepening markets. The problem that has often been raised since the 1990s is that much of what is identified as exclusion, especially in developing countries, might in fact represent impoverished or even exploitative forms of inclusion, or inclusion on poor terms, as synthesised by Gore and Figueiredo (1997). DuToit (2004) coined this as adverse incorporation and Du Toit and Hickey re-coined it as adverse incorporation and social exclusion. However, the idea is as old as theories about capitalism itself, whereby the exploitation of labour is understood to occur through their inclusion into capitalist processes, not their exclusion. Li (2007) nonetheless raises the spectre that contemporary capitalism might be evolving towards greater degrees of exclusion than in the past through the creation of ‘surplus populations’, although the question again is whether she is simply referring to the widespread informalisation of labour in developing countries that, as in the past, plays a functional role within capitalism of keeping wages low and workers disciplined.

To the extent that the problem is with the terms of inclusion and not with exclusion per se, simplistic promotions of inclusion might actually make matters worse. This is perhaps best exemplified by programmes of financial inclusion for the poor that are in effect predicated on quite exploitative terms (e.g., see Dymski 2005). Examples include the subprime crisis in the US (Dymski 2010; Dymski et al. 2011) or else many of the varieties of microfinance currently in practice, as discussed by Bateman (2010, 2012), Roy (2010) and Mader (2015, 2018), as well as in a compilation of articles edited by Bateman (2017).

These criticisms more or less accept the location of social exclusion within the space of poverty but then challenge whether notions of exclusion are necessarily the most appropriate to understand the cases of poverty that they are associated with. Another, subtler critique of the social exclusion approach questions its association with poverty and, in this sense, pushes back on the assertions made by Amartya Sen, as mentioned above. In other words, the concept might be redundant according to the ways it is conventionally defined and operationalised, but it nonetheless remains popular in both policy and academic discourse precisely because it captures the idea of something that is not quite the same as poverty. It is like an itch that we keep calling poverty but often is not. Indeed, the potential of this realisation is that it helps to move poverty studies away from its obsession with thresholds and into broader considerations of social integration.

This critique of both the dominant conceptions and the dominant criticisms of social exclusion can be clarified by situations where what we call social exclusion does not overlap with poverty, however conceptualised and measured, or where such exclusions worsen with movements out of poverty, such as among relatively non-poor and/or upwardly mobile people. As discussed below, examples include international migrants who face subordinating exclusions in their attempts to integrate into their host countries, or university graduates from disadvantaged minorities who struggle to compete for employment in the upper strata of the labour market, in accordance with their levels of education, and yet who are not necessarily poor as a result. Both examples would not necessarily fall under the rubric of capability or relative poverty (unless, of course, we tautologically define an exclusion as a capability failure, which is a danger of the capability approach, as discussed in the previous chapter). They might not even be reflected as rising inequality. In such cases, it would be wrong to assume that social exclusion is the same as poverty, whether absolute or relative.

This critique is also clarified by what I refer to as causal and positional relativity (Fischer 2008a, 2011a). The causal refers to the idea that there are multiple often contradictory processes at work within any condition or state of being, such that a person might face both exclusionary and inclusionary processes at the same time. Some of this might be conceived in terms of multidimensionality, such as exclusion from health insurance but not from labour markets, but not necessarily all, given that these multiple causalities can even occur within a single dimension. As a result, the identification of exclusion in these situations depends on which causal process we decide to focus on. It is not necessarily possible to essentialise the characteristics of this exclusion onto an overall identity or condition of the person or group in question. This point is similar to the often-noted insight that exclusion is usually predicated on patterns of inclusion, or vice versa, as referred to in Gore and Figueiredo (1997). It is not exactly the same as the adverse incorporation point made by Du Toit (2004), insofar as it recognises characteristics of exclusion as distinct from inclusion, operating together rather than being conflated. As such, it bears more similarity with some of the French sociological scholarship (e.g., Roulleau-Berger 1999), as discussed further below.

Positional relativity means that what we call exclusion is relative to a person’s position in a social hierarchy, both objectively and subjectively. In other words, a person can be an excludee and an excluder at the same time and will tend to perceive their exclusion (or inclusion) relative to the social strata immediately surrounding them. This is analogous to insights from the wellbeing literature that one’s subjective perception of wellbeing depends in part on one’s position or situation relative to one’s immediate surrounding, which serves as a basis of social comparison.² It also ties into feminist notions of intersectionality in that positional relativity can apply across many dimensions of comparison, such as ethnicity, class, caste, clan, occupations, gender, generation or location.

In these respects, various poverty approaches are only capable of reflecting what we call exclusions operating at the bottom of a social hierarchy. They do not capture similar processes operating higher up in a social hierarchy among non-poor people, even though these dynamics might be crucial for understanding processes of social stratification, marginalisation, disadvantage and discrimination, all of which can occur in the absence of poverty. Overlaps obviously occur between these processes and poverty, which merit our urgent attention, although we render a disservice to our ability to understand and address the interactions if we merge concepts together.

Differentiating poverty from what we call social exclusion is also important from a policy perspective. As has been noted with respect to gender, such as by Jackson (1996), policies that are designed to address poverty might not be appropriate to address things such as disadvantage or discrimination. Indeed, the policy objectives of addressing each might be quite distinct, such as when addressing ethnic disadvantages that do not necessarily overlap neatly with socio-economic disparities.

If differentiated from poverty, the social exclusion approach might provide additional analytical insight to understand these processes that overlap but also extend beyond the space of poverty. However, the question then is that, for all of the vagueness, confusion and politically loaded insinuations associated with the concept, why not jettison it altogether? How does the concept of social exclusion add to our understanding of processes of stratification, subordination and segregation, or of disadvantage and discrimination, more than we are already able to obtain through these existing sociological concepts? This has been debated intensely in the literature, but with no convincing or conclusive results. In effect, the enticement of the concept is that it permits a discrete consideration of social power relations – the ways in which social power relations work to confer or deny access to resources or opportunities – with a jargon that does not cause alarm to the establishment. But, then, what are the costs of trying to assuage the establishment, such as the cooptation of the concept into market inclusion agendas, as noted above, versus calling a spade and spade and being done with it? The danger, as usual, is in replacing existing, established and relatively clear social-scientific concepts and methods with vague and unspecified new ones.

These arguments are made in four sections. The first offers a brief introduction to the concept of social exclusion and the second synthesises the many ambiguities surrounding the concept. The third elaborates on the point about differentiating what we have come to call social exclusion from poverty and the fourth provides some practical examples of why it is important to do so. The conclusion returns to the fundamental point that other, more precise social science concepts are arguably better-suited for the purpose of studying and addressing the range of issues that have been stuffed under the social exclusion umbrella, and with clearer political and policy implications.

A synthesis of the social exclusion approach

In the canonical accounts of the social exclusion approach, the concept is generally said to have been coined by Réné Lenoir (1974) in his seminal book, Les Exclus. His work essentially referred to the increasing inability of European welfare states (and, in particular, that of France) to deal with cases of people falling through the cracks of the welfare system because they were not able to work. Work was crucial because these systems were designed on the basis of full employment, whereby at least one member of every family would establish the basis for welfare entitlements through regular, full-time, formal employment. Lenoir identified those who were unable to work for either physical or mental reasons as the excluded, who accounted for a minor but significant and growing proportion of the population.

The idea was then increasingly taken up in 1970s and 1980s in Europe, although with somewhat different interpretations. Instead of dealing with the marginals who could not work, it was increasingly applied to the emergence of chronic, long-term, involuntary unemployment, in the context of economic crisis and restructuring and the increasingly inability of the welfare states to deal with the rising trend of people who were able to work but could not find work. This interpretation of social exclusion then became increasingly mainstreamed in the EU and the UK in the 1990s, as discussed further below.

The International Institute of Labour Studies (IILS) of the International Labour Office (ILO) then launched a research project on social exclusion in 1993 with the support of the UNDP. The aim was to refashion the concept from its roots in European social policy discourse to a wider, less Eurocentric global application, for application to anti-poverty policy in particular (see Rodgers et al. 1995, p. vi). Attempts were made to re-interpret the concept from a way of describing processes of marginalisation and deprivation in rich countries with comprehensive welfare systems and where the vast majority of the workforce is integrated into formal employment, to developing countries where universal welfare provisioning is mostly absent and formal employment usually only covers a small minority of the workforce. The project was intended to contribute to the discussion at the World Summit for Social Development in 1995 and to explore ways in which the analysis of exclusion could make anti-poverty strategies more effective (ibid).

This need for a more general and generic refashioning was seen as particularly urgent given the economic crises and structural adjustments of the 1980s and early 1990s. These were undermining whatever limited welfare developments had previously been achieved by developing countries, especially in Latin America. Crises and neoliberal responses were also associated with an intensification of informality rather than its attenuation, despite neoliberal arguments that deregulation would facilitate formalisation, as argued, for instance, by De Soto (1989). The fact that structural adjustments were being guided by the same overarching ideology, with comparable dynamics of increasing precarity and insecurity as those occurring in northern welfare states, lent weight to the project of extending the concept of social exclusion to developing countries even despite the absence of similar institutional settings.

The effort to render the concept less Eurocentric was nonetheless criticised as Eurocentric, from the perspective of trying to impose concepts that derive from a European context onto non-European contexts. The criticism is that there is little to be gained from applying a concept formulated in rich post-industrial societies with already-existing comprehensive formal labour markets and welfare systems to poor countries where, if the same metres of identification and measurement were to be used, the majority of the population would be deemed ‘excluded’.³ However, this criticism can be easily dealt with by loosening the institutional specifications of the concept and treating it more generically in terms of the principle that patterns of exclusion are relative to whatever prevailing modes of inclusion are operating in any particular context, whichever the context. This point is similar to arguments made by Hillary Silver (1995), as discussed further below.

The Eurocentric critique can also be turned on its head by simply reversing the charge and noting that the intellectual origins of the concept might have plausibly derived from the South in any case. We need only recall that Latin American structuralist and dependency scholars were using terms that are very analogous to exclusion since the 1950s and 1960s, such as marginalisation and peripheralisation, if indeed they were not already referring directly to the term ‘exclusion’ itself. As noted by Gore (1995, p. 4):

it has been suggested that the emerging social problems of Europe reflect a process of ‘Latin Americanization’, in the sense that European economies and societies are moving closer in their forms of organization to those of Latin America … As that occurs, the language to describe and analyse the situation in Europe is catching up to one already widely deployed in Latin America, where debates about marginalization were already vigorous in the 1960s.

In terms of lineages of thought, it is also notable that the Brazilian economist Celso Furtado, one of the foremost from the Latin American structuralist school, taught at the Sorbonne from 1965 to 1985, along with many other Latin Americans exiled from their countries during that time (Kay 2005, p. 1204). Furtado would have undoubtedly influenced his French colleagues. The presence in Paris during the 1950s and 1960s of other well-known Third World scholars using similar terminologies is also noteworthy, such as Samir Amin (e.g., 1968, 1970, 1973). Likewise, René Lenoir spent several years as a technical advisor to African governments prior to his 1974 publication, further reinforcing this hypothesis that his ideas might have been the product of creative intellectual fertilisation from South to North.

Irrespective of these debates in the history of thought, the literature has generally settled around a number of conventional definitions of social exclusion. In the IILS project to universalise the concept, Rodgers (1995, p. 44) defined it quite generally and vaguely ‘as a way of analysing how or why individuals and groups fail to have access to or benefit from the possibilities offered by societies or economies.’ Meanwhile, the way the concept has been defined in the UK and EU policy and academic literature is more in terms of the social aspects of deprivation, such as the isolation that one presumably experiences when one is (relatively) poor, or else how experiences like isolation might cause or reinforce relative poverty (e.g., see the volume edited by Room 1995b). The latter approach is decidedly individual in its identification of attributes, in the sense that the attributes of exclusion are not necessarily based on identity, even though identity could be one cause of exclusion. This is in contrast to some of the approaches of applying the concept to developing countries, which tend to define social exclusion in terms of group-based discriminations, as discussed further below.

The exercise of determining definitions obviously raises a variety of questions. From what should people be considered excluded? Or by what or by whom? Some argue that social exclusion differs from poverty because it begs these types of questions and points towards non-material aspects of poverty, such as status or denial of access, or else highlights concepts such as disadvantage, discrimination, marginalisation and the relational processes that are related to these phenomena. It also highlights the importance of citizenship rights and the importance of institutional analysis in understanding patterns of inclusion, in terms of how formal or informal rules and norms are formed and practised to segregate, stratify, structure and order populations, such that some experience restrictions to the rights enjoyed by others, whether for reasons of identity, property, rights, etc.

In terms of methods, various approaches have also been proposed. In the UK and EU approach mentioned above, the main efforts have been to quantify various social dimensions of poverty. Others propose measures of group representation, such as proportions of a group in particular positions or experiencing particular outcomes versus their population share, such as the percentage of African-Americans in the US prison population versus their population share. Of course, the latter measures do not detect exclusion per se. Rather, they are measures of representation aimed at revealing discrimination, which might or might not result from exclusionary practices. Other measures include legal approaches that attempt to identify legally defined exclusions, or sociological analyses of daily practices, in cases where formal legal provisions are inclusive but practices and experiences are not.

Much like the capability approach, one of the main challenges of measuring social exclusion is to actually identify exclusions rather than the spaces in which we might expect exclusions to be occurring, which will be discussed further below. Moreover, the measures that identify discriminations are group-based measures and, much like the challenges facing the HDI discussed in the previous chapter, the challenge is to design measures that are able to identity individual experiences of exclusion.

While there has been some development on individual-based measures, particularly with respect to the UK/EU work on social exclusion as noted above, the problem is that the proposed and adopted individual-based measures are essentially measures of poverty rather than necessarily representing exclusion. For instance, Burchardt et al. (1999) propose a notion of participation in five types of activity, including consumption, savings, production, political and social. However, this essentially boils down to income and asset poverty, and unemployment, with the addition of social isolation and lack of political participation. (The inclusion of political participation is odd and normative, in the sense that there might be many otherwise ‘normally’ integrated people who are not members of political parties or campaign groups, or who do not vote.)

This constantly leads back to the central criticism in the literature, which is that the social exclusion approach is duplicating or even distracting attention away from poverty, that it is redundant with poverty concepts or that it leads to a different identification of the poor or different anti-poverty policies. Indeed, this last point is a conclusion from Burchardt et al. (ibid.), who identified only a few people in their study who were excluded on all dimensions in any one year and even fewer who experienced multiple exclusions for the whole period, although of course this is not surprising given their inclusion of the political participation variable, as noted above. They argue that this ‘supports the view that treating different dimensions of exclusion separately is preferable to thinking about social exclusion in terms of one homogeneous group.’ While this is a sensible point, it still begs the question of what, exactly, is different between this approach and multidimensional measures of poverty, as discussed in the previous chapter.

The ambiguities of social exclusion

In order to expand on the last points above, it is useful to clarify three closely related problems in the literature on social exclusion since the early 1990s that result in strong degrees of ambiguity. One is that most authors invariably operationalise social exclusion as a static description of outcomes or a state of being, even though most also agree that the strength of the concept lies in the attention it brings to dynamic processes. Second, as noted above, there is a propensity to treat the concept of social exclusion as a type of poverty, despite the fact that it is often recognised that exclusion can occur in the absence of poverty. The third problem derives from the ambiguous use of the terms ‘relative’ and ‘relational’.

Processes and states

A good starting point to demonstrate the ambiguous reference to processes and states of being is found in the IILS work in the mid-1990s. Most scholars in this project came to agree that the value-added of the social exclusion approach, over other concepts of poverty or deprivation, is its focus on processes, particularly social processes such as disadvantage and/or impoverishment. For instance, in the introduction of their summary of one key debate organised by the IILS, in which they define exclusion as both a situation or a process of marginalisation or the fragmentation of social relations, Gore and Figueiredo (1997, p. v) argue that the concept focuses attention on processes that lead to disadvantage, impoverishment or illbeing, rather than an identification of excluded individuals or groups in an absolute sense. Similarly, in their comprehensive review of four main ways of defining and measuring poverty, Laderchi et al. (2003) conclude that a key strength of the social exclusion approach, in comparison to the monetary, capability or participatory approaches, is that it ‘is the only one that focuses intrinsically, rather than as an add-on, on the processes and dynamics that allow deprivation to arise and persist’ (p. 260). This position is also reflected in much of the UK-oriented scholarship, such as that by Berghman (1995), who argues that social exclusion needs to be seen as a dynamic process and not a state or outcome. Peter Townsend (2002, p. 7) makes a similar point, that inequality and poverty correspond to an idea of state, whereas polarisation and exclusion correspond to an idea of process.

However, in their attempts to operationalise social exclusion, most scholars invariably end up treating it as a state, outcome or condition of being, rather than as a process. This tendency derives from the unbearable itch to apply thresholds, reproducing the equivalent of poverty lines, as an effort to identify who, exactly, are the excluded, and below what threshold they should be considered excluded. The resultant search for an appropriate set of outcome indicators invariably leads to a static conceptualisation of social exclusion, which is arguably in tension with the emphasis on process. In particular, the outcome indicators mostly do not identify processes but instead the outcomes that are presumed to result from exclusion or that perhaps cause exclusion (such as income or asset poverty, as noted above). This also leads to dangers of tautology whereby causal processes are defined by their results.

Moreover, the outcome-based operationalisation of social exclusion also encourages a methodology-driven influence on the definition of social exclusion. As noted by Levitas (2006) with respect to the work on social exclusion in the UK and EU, ‘the necessity of multiple indicators means that it is possible to draw up a provisional set without clarifying underlying definitions and relationships, and without any statement of priorities’ (p. 127). In other words, operationalisation sweeps the definitional problems under the carpet rather than solving them. It is therefore no wonder that scholars have had difficulty in determining the applicability of social exclusion outside a European context, given that there is no clarity or precision on the concept even within Europe.

This problem of slipping into static notions is similar to the problem inherent in the capability approach of measuring potentials in terms of functionings, that is, outcomes such as life expectancy, morbidity, literacy and nutrition levels. As noted in the previous chapter, this removes the key philosophical insight of the capability approach on freedom and essentially renders it the same as the basic needs approach it set out to supplant. Social exclusion comes at the problem from a different angle – the measurement of processes rather than potentials – although it ironically ends up in more or less the same predicament of being measured according to various static outcome indicators of functionings or basic needs.

As a result, many definitions of social exclusion lack clarity as to whether social exclusion refers to a state or a process. Most authors compromise by treating it as both. Gore and Figueiredo (1997) accept that the dimensions of state and process overlap and that this is not necessarily contradictory. They suggest that (p. 18):

the former offers a way of describing social exclusion and can for instance be used to define situations of permanent exclusion. In the latter, the focus is on the mechanisms which create or recreate exclusion, and on how poverty and deprivation are associated with structural economic and social change.

Laderchi et al. (2003) summarise this position by stipulating that ‘the definition of [social exclusion] typically includes the process of becoming poor as well as some outcomes of deprivation’ (p. 258). Similarly, Beall and Piron (2005) offer an abbreviated working definition of social exclusion as ‘a process and a state [deriving from exclusionary relationships based on power] that prevents individuals or groups from full participation in social, economic and political life and from asserting their rights’ (p. 9). Besides the recurrent tautology of defining exclusion as exclusion in these attempts at definition, they seek to solve the ambiguity between state and processes by grounding social exclusion in an understanding of poverty.

Social exclusion and poverty

The tension between states versus outcomes is closely intertwined with the common association of exclusion with poverty. As in the example of Burchardt et al. (1999) mentioned above, the definition of social exclusion is usually collapsed into a conception of poverty, either as the aspects of social deprivation involved in multidimensional concepts of poverty, or else as social processes leading to poverty. In other words, many definitions imply that if exclusion does not lead to some impoverishment, it is therefore not exclusion, or else it is not one that warrants our attention, and that it is mostly in the space of overlap where we should focus our concerns. For instance, in their efforts to resolve the debate on how best to differentiate social exclusion from poverty, Gore and Figueiredo (1997, p. 10) argue that the former refers to processes of impoverishment. ‘Its value then is that it enables causal analysis of various paths into and out of poverty, getting beyond the unhelpful lumping together of diverse categories of people as “the poor.”’ Similarly, Laderchi et al. (2003) explicitly treat social exclusion as one of four main ways of defining and measuring poverty.

Gore (1995, pp. 1–2) nonetheless clarifies that the early French debates of the 1970s and 1980s did not necessarily equate social exclusion with poverty as such, but with processes of social disintegration. The concept only became more closely equated with poverty in the early 1990s after the European Commission defined social exclusion in relation to a certain basic standard of living. In the UK, Levitas (2006, p. 126) notes that the distinction between social exclusion and poverty is ‘sometimes masked by references to “poverty and social exclusion” as an inseparable dyad’. Pantazis et al. (2006, p. 8) also note that the 1999 Poverty and Social Exclusion Survey for the UK proposed impoverishment as one of four dimensions for measuring social exclusion, alongside labour market exclusion, service exclusion and exclusion from social relations. This selection thereby includes a notion of poverty into its implicit definition of social exclusion. Levitas (2006, pp. 130–131) remarks the same with the indicators adopted by the EU Social Protection Committee in 2001, which included income and labour-market position as well as deprivations in education, housing and health. The EU position has basically not changed since.⁴

In the development studies literature, such as Laderchi et al. (2003), social exclusion is usually defined as poverty or as processes leading to poverty. For instance, all of the working definitions in the series of country studies commissioned by the IILS in the 1990s essentially treat social exclusion as contextually defined forms of relative or capability poverty – see Rodgers et al. (1995) for detailed presentations and a summary by Gore and Figueiredo (1997, pp. 17–18). The India study defined social exclusion as a denial of the basic welfare rights that provide citizens positive freedoms, and Appasamy et al. (1995, p. 238) define these as basic needs in education, health, water and sanitation, and social security. The Thailand study defined it as a non-recognition or disrespect for the citizenship rights on which livelihood and living standards depend. The Russia study defined it as material deprivation and infringement of social rights, defined mainly in terms of employment. The Tanzania study defined it as both a state and process, with the state being equivalent to relative deprivation, while the processes were socially determined impediments to access resources, social goods or institutions. The Yemen study defined it as social segregation, where some individuals and groups are not recognised as full and equal members of society. The Peru study defined it in terms of the inability to participate in aspects of social life considered important (Gore and Figueiredo, 1997, pp. 17–18). Figueroa et al. (1996) deemed that its analytical value comes from its elucidation of social processes which contribute to social inequality (p. 201). The general definitional gist remains consistent throughout; all these studies treat social exclusion as either relative or capability deprivation.

Subsequent elaborations of the concept in development studies scholarship continued along these lines. In a workshop convened by the Institute of Development Studies in Sussex in 1997, De Haan thought that despite its overlap with the concept of poverty, the concept of social exclusion could be useful nonetheless because it focuses on processes and because it is multidimensional in nature. However, he was doubtful whether these aspects made it different from poverty. The distinction was only clear if one adopted a very narrow view of poverty, whereas ‘much of the current debate on poverty, especially in developing countries, was concerned with wider concepts of relative deprivation, ill-being, vulnerability and capability’, and was thus very similar to the social exclusion approach (cited in O’Brien et al. 1997, p. 3). Bhalla and Lapeyre (1997) operationalise social exclusion by essentially focusing on depth of poverty and income inequality measures, and a variety of social aspects purporting to measure access but that essentially boil down to a measurement of various functionings and outcomes (pp. 425–426). Saith (2001) also explores the feasibility of operationalising the social exclusion concept in developing countries. She notes (p. 1) that efforts

appear to largely result in a repetition of research that has already been conducted within frameworks that have developed in developing countries (basic needs, capabilities, sustainable livelihoods, risk and vulnerability, participatory approaches) in parallel to the ‘social exclusion’ concept in industrialised countries.

In order to avoid the relabelling of poverty studies, she suggests that ‘rather than trying to transplant the concept, it might be worth concentrating on incorporating the advantages of “social exclusion” like its emphasis on process into existing frameworks in developing countries’ (p. 14). In all these examples, the authors struggle to break free from existing poverty approaches, but ultimately return to the fold.

This point is made by several authors in both the social policy and development studies literatures. Levitas (2006, p. 126) notes that even the aspects of social exclusion dealing with social relations, which are deemed to be among of its most important contributions, such as exclusion from social participation, were part of the earlier conceptualisation of relative deprivation by Townsend (1979). Room (1999) makes the same point, that the multidimensional, dynamic and community aspects often promoted as the novelties of the social exclusion approach all existed in the ‘classic’ studies on poverty. He suggests that the more original element of social exclusion is found in its emphasis of relational issues (inadequate social participation, lack of social integration, and lack of power), versus the primary emphasis of poverty on distributional issues (pp. 167–169). Bhalla and Lapeyre (1997, p. 417) contend with this distinction – made in earlier papers by Room, e.g., Room (1995a), arguing that the broad concept of poverty (i.e., relative and capability deprivations) covers both the distributional and relational aspects of deprivation referred to by Room (1994, 1995a), hence bringing even these elements back into the poverty fold.

Amartya Sen (2000) makes exactly this point in his own rendition of social exclusion. He argues that, by way of relationality, social exclusion constitutively describes one aspect of capability deprivation and instrumentally causes further diverse capability failures (p. 5). In other words, he paraphrases in the negative what he says about freedom in Development as Freedom (1999): the relational features of social exclusion ‘enrich the broad approach of seeing poverty as the lack of freedom to do certain valuable things’ (ibid.). He notes that this emphasis is nothing new and refers back to Adam Smith’s concern with ‘deprivation in the form of exclusion from social interaction, such as appearing in public freely, or – more generally – taking part in the life of the community’ (p. 7).

However, by reducing the social exclusion approach into an analysis of relationality, Sen argues that its helpfulness does not lie in its conceptual newness, but in the attention it focuses on the role of relational features in deprivation (p. 8). For this reason, if the language of exclusion is to add value to our understanding of deprivation, ‘it is crucial to ask whether a relational deprivation has been responsible for a particular case of [deprivation]’ (pp. 9–10). He therefore concludes that, if understood in this way, the ‘perspective of social exclusion reinforces – rather than competes with – the understanding of poverty as capability deprivation’ (p. 46). In other words, he effectively relegates the concept of social exclusion to an adjunct position within his own theoretical agenda.

Unfortunately, Sen seems possessed with establishing the authority of his own capability approach rather than with distinguishing the social exclusion approach in its own right, to the extent that he almost appears defensive. For instance, in his rendition of social exclusion, Sen contends with an implicit criticism of the capability approach made by Gore (1995, p. 9), arguing that there is nothing in the capability approach that would doom it to be excessively individualist and insufficiently social, and thereby causing it to miss a focus on the relational features of deprivation (Sen 2000, p. 8). Indeed, it is ironic that many of the efforts to clarify or elaborate a social exclusion approach, such as the IILS studies, have instead ended out adding operational rigour to the capability approach. Conversely, Sen avoids the concept of relative deprivation, even though it might seem better-suited than the capability approach for dealing with many of the concepts that he discusses in relation to social exclusion, such as the inability to participate in a community due to a lack of means relative to social norms. In any case, he offers no guidance on how to differentiate social exclusion from poverty.

Some authors in the development studies scholarship attempt to avoid an explicit association between social exclusion and poverty but nonetheless use an idea of norms to situate exclusion within a social context. Examples include the reference to ‘full participation’ mentioned above (cf. Laderchi et al. 2003, p. 257; Beall and Piron 2005, p. 9), or ‘normal activities’ (Stewart et al. 2006, p. 4). These references are probably influenced by the EU definition of social exclusion in the mid-1990s as a ‘process through which individuals or groups are wholly or partially excluded from full participation in the society in which they live’ (Laderchi et al. 2003, p. 257). Here again, exclusion is tautologically defined as exclusion and, despite the effort to explicitly define it without reference to poverty, the understanding of ‘full participation’ or ‘normal activities’ places the definition into a metre of relative deprivation without explicitly stating this as such.

Similar attempts have been made to refine the concept of social exclusion along the lines of rights or choice. However, once the implications of such an approach become specified, it again becomes difficult to see how it differs in practice from either the capability or relative deprivation approaches, particularly given the emphasis on freedom in the former and the emphasis on norms in the latter. For instance, Schulte (2002) deals with social exclusion within a rights framework, treating it as the denial of a whole range of rights denoted by the concept of social citizenship. He stipulates that these include ‘the right to social security and economic wellbeing, to the right to a full share in the social heritage and to the life of a civilised human being according to the normal standards prevailing in that society’ (p. 121). In other words, once he stipulates the meaning of social citizenship rights, his approach essentially seems to be a reformulation of the capability approach, with an emphasis on capabilities defined by the relative norms of a society.

Choice was also emphasised in the adoption of an initial definition of social exclusion by the Centre for the Analysis of Social Exclusion (CASE): ‘an individual is socially-excluded if he or she does not participate in key activities of the society in which he or she lives; … the individual is not participating for reasons beyond his/her control; and he or she would like to participate’ (Burchardt et al. 2002, pp. 30 and 32). However, it is again not clear how this emphasis on lack of choice (or un-freedom) is simply not a restatement of Sen’s position on capability failure, as discussed below. In any case, Levitas notes that once operationalised, this CASE approach becomes limited to the first of their clauses and sidesteps the issue of choice for pragmatic reasons (2006, p. 134). In other words, as noted above, this approach also falls into the same predicament as the capability approach of reducing the measurement of potentials to that of functionings. The definition also runs into problems of determining which ‘activities’ should be considered ‘key’, in particular because such considerations are hugely related to one’s social position, as discussed further below.

Kabeer (2006) makes a notable attempt to differentiate social exclusion from poverty by treating social exclusion as an analysis of processes of disadvantage, although she does this exclusively through the lens of identity discrimination. She elaborates (p. 3) that it

reflects the multiple and overlapping nature of the disadvantages experienced by certain groups and categories of the population, with social identity as the central axis of their exclusion. It is thus a group or collective phenomena rather than an individual one.

Thus, she contends (p. 2) that

understanding the social dimensions of inequality, hitherto ignored in mainstream poverty analysis, provides new lens through which to view the issue of chronic disadvantage … These revolve around social identity and reflect the cultural devaluation of people based on who they are (or rather who they are perceived to be).

In this sense, Kabeer frames the ‘social’ of social exclusion in terms of social groups and identities, sharing much in common with the recent scholarship on intersectionalities. This is in contrast to an understanding of the ‘social’ in most of the UK- or EU-oriented literature, which refers to the social aspects of deprivation, which can operate at the individual level and without the intermediation of identity-based discrimination. Thus, while her approach makes sense within the South Asian context, at least in terms of understanding one aspect of disadvantage, it is hardly encompassing of the broader dimensions of exclusion, particularly in terms of how the concept came to be used in the 1980s with reference to economic restructuring.

Furthermore, her treatment of disadvantage is explicitly connected to an analysis of poverty. She notes in the beginning that the ‘durable nature of this form of disadvantage means that people bearing devalued identities are likely to be disproportionately represented among the poor, as well as among the chronic poor’ (ibid., p. 2). The rest of her paper therefore looks at social exclusion mostly from the perspective of an enriched processual understanding of the identity-based dynamics of poverty. Thus, it does not as such resolve the ambiguity between social exclusion and a wider multidimensional understanding of poverty, but implicitly accepts the fact that exclusions are only worthy of our attention if they overlap with poverty.

It is not clear, however, why she asserts that these aspects have been ignored in mainstream poverty analysis, given that her approach more or less echoes certain strands of the IILS publications ten years previously. For instance, in the debate summarised in Gore and Figueiredo (1997), there were several suggestions to see social exclusion as the generation of multiple disadvantages. Citing the work of ‘Waltzer’ (perhaps Michael Walzer?), these can then further evolve into ‘destructive synergies’ or ‘radical disadvantage’ (p. 43). It was further suggested that social exclusion be understood as a ‘second-tier’ concept of risk regime that expresses the interaction between social risk factors (such as unemployment, lack of access to social services or family breakdown), the cumulation of these factors and the diminished capacity of groups to respond to risk.

The notion of social exclusion is therefore not a broad notion of poverty which encompasses non-material aspects and deprivations which do not arise from lack of resources. Rather poverty is a part of social exclusion. It is one of various risk factors which together cumulate to form a risk regime. (Gore and Figueiredo 1997, pp. 40–41)

While the source of this perspective was not identified, it resembles arguments made by Room (1999) that the element of catastrophic discontinuity in relationships with the rest of society offers the most essential contribution of social exclusion, in combination with its emphasis on relational elements. He argues (p. 171) that

to use the notion of social exclusion carries the implication we are speaking of people who are suffering from such a degree of multi-dimensional disadvantage, of such duration, and reinforced by such material and cultural degradation of the neighbourhoods in which they live, that their relational links with the wider society are ruptured to a degree which is to some considerable degree irreversible. We may sometimes choose to use the notion of social exclusion in a more general sense than this: but here is its core.

This perspective certainly provides perhaps one of the most convincing distinctions of social exclusion from poverty, although it still ends up treating social exclusion as a catastrophic outcome occurring horizontally at the bottom of a social hierarchy, parallel to poverty.

It also brings us back to the idea of thresholds, in terms of being able to social-scientifically identify a point below which things start to fall apart, as discussed in previous chapters – ironically because the title of Room’s 1995 publication is Beyond the Threshold (Room, 1995b). Indeed, he argues that it may be better to use the notion this way for it to be ‘useful as an analytical concept and as a point of reference for policy design, rather than to use “social exclusion” as no more than a synonym of “disadvantage”’ (ibid.). However, his distinction along these lines is not made through precise definition, but rather through a complex consolidation and integration of five elements, most of which already exist in the classic studies of poverty, as he consistently argues. In other words, his own argument that the other aspects of exclusion can be dealt with through a multidimensional approach to understanding poverty could equally apply to his conclusion on catastrophic discontinuity.

Relativity and relationality

The inconsistent meanings implied by the use of the terms ‘relativity’ and ‘relationality’ constitute a third problem in the literature, which in turn reinforces the ambiguity between social exclusion and poverty. Relativity is typically used in two closely related ways; relative poverty (exclusion relative to social norms) and contextual relativity (exclusion depending on societal modes of integration or incorporation). Laderchi et al. (2003, p. 258) draw a close connection between these two meanings, in that norms are determined by context.

The latter, contextual meaning of relativity owes much to the work of Hilary Silver (e.g., 1995), who proposes a threefold typology of the multiple meanings of exclusion inspired by the three models of welfare capitalism elaborated by Esping-Anderson (1990). These are situated in three different theoretical perspectives, political ideologies and national discourses (the solidarity, specialisation and monopoly paradigms). Without going into any detail on these paradigms (in part because their substantive content refers mostly to OECD countries), it suffices to note that Silver purposely avoids offering a definition of social exclusion precisely because she sees the ambiguity as offering a window of opportunity through which to view conflicting social-science paradigms and political ideologies. ‘This is because at the heart of the question “exclusion from what?” is a more basic one, the “problem of social order” under conditions of profound social change’ (Silver 1995, p. 61). Even though her three types refer mostly to European contexts, her work has nonetheless helped to extract the concept of social exclusion out of an association with a specific context and to theorise it in more generic terms.

Her stance in turn inspired the approach of the IILS, particularly considering that she was a central player in its initiatives of the 1990s. Thus, Gore argues that

a precise definition of social exclusion depends on the paradigms of social integration and citizenship and the cultural environment prevailing in a society. These structure people’s sense of belonging and membership and consequently the perception of what is exclusion and inclusion in their society. (Gore 1995, p. 8)

A similar line is taken up in the work of Atkinson (1998), who emphasises relativity as one of the three main characteristics of social exclusion, alongside agency and dynamics. Relativity in this sense refers to the fact that the meaning of exclusion is relative to a particular society, and Atkinson argues that this is an important element that differentiates social exclusion from the concept of poverty (pp. 13–14). Similarly, Laderchi et al. (2003) note that definitional problems ‘are especially great in applying the concept to developing countries because “normality” is particularly difficult to define in multipolar societies, and because there can be a conflict between what is normal and what is desirable’ (p. 259). Similar to Silver, this sense of relativity is generally seen as one of the principle strengths of the social exclusion approach.

Silver does raise some important interpretative issues, although her avoidance of definition by way of typology runs the risk of giving licence to use the concept as an ad hoc descriptor of any variety of multidimensional deprivation or disadvantage. Or else, it risks rendering the concept into an entirely descriptive template and aborting the analytical project of understanding exclusion as a causal process in its own right. Moreover, the typological approach also avoids the ambiguities between exclusion and poverty, which are not necessarily the result of relativity, but rather of conceptual imprecision. Indeed, as noted above, the allowance for definitional plurality was so loose in the series of IILS studies that that the concept came to mean just about anything to anyone, so long as it generally referred to some negative sense of multidimensional deprivation or disadvantage. Thus, while the various elaborations of social exclusion represent very valuable deepening of our understanding and measurement of deprivation, they do not solve the conceptual problem of differentiating social exclusion from poverty.

On a similar note, there is inconsistency in the literature on the exact meaning of the term ‘relational’ or ‘relationality’. As argued by Room (1999), a relational deprivation need not imply an intentional act of exclusion, but merely a breakdown in social relations due to some deprivation, or else that social isolation leads to other deprivations, such as when a lack of social integration leads to poor health or education. In this sense, ‘relational’ refers to the absence of social relations, not necessarily that exclusion is caused by a relation between people.

However, others such as Beall and Piron (2005, p. 11) refer to the term ‘relational’ as an intended act of exclusion by an excluder towards an excludee. This latter usage, which is quite common, is actually closer to the term ‘agency’ used by Atkinson (1998), as also emphasised by Laderchi et al. (2003). Sen (2000, pp. 14–18) is ambiguous on this point and uses the two meanings interchangeably, noting that ‘relationality’ is already dealt with in the capability approach. He also refers to ‘active’ or ‘passive’ exclusion, which seems to be his way of dealing with agency. Regardless, we have to question whether exclusion necessarily needs to be intentional, insofar as impersonal structural dynamics can equally produce exclusion despite the best of relational intentions, as in the case of neighbourhood gentrification due to rising house prices. Indeed, the earlier conceptions of social exclusion developed in the 1970s and 1980s were precisely in response to such structural processes of social disintegration related to economic restructuring. Any attempt to define the concept should therefore engage with these ‘non-relational’ structural dimensions. Nonetheless, even if we accept ‘relationality’ or ‘agency’ as distinguishing elements of social exclusion, these arguably also play an important role in the capability approach, as noted earlier.

These various meanings of relativity and relationality throw a serious wrench into the definition or identification of exclusion. In particular, it is very difficult to determine norms within highly segmented, heterogeneous and hierarchical settings, where one person’s norm is completely different from another’s, or is at least perceived to be and accepted as such. However, this perspective also offers a way out of the ambiguity of social exclusion by implying that exclusions (or what we call exclusions) occur vertically across social orders, whereas poverty occurs horizontally at the lower end of social orders. In other words, all individuals or groups can perceive or experience an exclusionary process from whatever their position.

Differentiating social exclusion from poverty

These ambiguities in the scholarship dealing with social exclusion are substantive and not merely semantic. Indeed, ambiguity is a more serious criticism than the charge of Eurocentrism in the application of social exclusion to developing countries. As noted above, the Eurocentric critique can be dealt with by loosening the institutional specifications of the concept, or by reversing the charge and noting that the intellectual origins of the concept might have derived from the Global South in any case. However, the same cannot be said for the ambiguous association of social exclusion with poverty, which applies to both South and North. If the difference is merely semantic, as claimed by Sen (2000), then the whole project of trying to establish social exclusion as more than simply an adjunct way of describing various social aspects of poverty is put into question.

If merely semantic, then the often-noted criticism that the concept of social exclusion deflects political attention away from poverty and inequality would be particularly damning, although this criticism is more oriented towards its usage in the UK and Europe. For the rest of the world, the more significant implication is that, as noted in the Introduction, the association with poverty implies that the solution to poverty is to intensify inclusion. Indeed, this partly explains why the concept has been adopted so enthusiastically by the mainstream of development policy, such as by the World Bank.

The problem with attributing the solution to poverty as one of inclusion is that most cases of poverty in developing countries are better described as arising from the manner by which people are already included (i.e., low wage rates and poor working conditions) rather than exclusion (i.e., unemployment), as noted by a variety of scholars (e.g., Gore and Figueiredo 1997, pp. 41–42; Room 1999, p. 171; Du Toit 2004; Hickey and Du Toit 2007). Policies of inclusion that do not address these existing terms of inclusion might simply end up exacerbating exploitation, particularly when they end up facilitating such practices through liberalisation and deregulation. In contrast, people in such circumstances might well regard their situation as better when they are ‘excluded’, as highlighted in a study by Beall (2002) of municipal sweepers in Faisalabad, Pakistan, who used their identity-based social exclusion to secure livelihoods. Interestingly, however, the World Bank has recently come to define social inclusion as improving the terms by which individuals and groups can take part in society, particularly with regard to identity disadvantages, perhaps reflecting some influence from these previous contentions.⁵ Of course, the devil is in the detail, with respect to how the Bank defines ‘terms’ and ‘participation’ within such revised conceptions of inclusion.

It is from this perspective that Byrne (2005, p. 60) warns that the ‘babble – no other word is strong enough – by political elites about exclusion can serve as a kind of linguistic trick’, on one hand presuming a continued commitment to the values of social democracy, while on the other hand supporting globalisation and neoliberalism. For similar reasons, Clert (1999, p. 195) warns that the coexistence of different ways of using social exclusion discourses can serve to obscure policy orientations and generate false consensus. Indeed, the typical emphasis of agency and relationality can also lend weight to the tendency within mainstream development policy to focus exclusively on the failures and abuses of domestic policies and domestic elites rather than, for instance, the economic austerities imposed by structural adjustment programmes or by international economic integration.⁶

These criticisms would be potentially allayed if we could distinguish between social exclusion (or what we consider to be exclusion) and poverty, or if we can identify exclusions that do not lead to poverty and vice versa. This is recognised by many authors, such as Gore and Figueiredo (1997) or O’Brien et al. (1997), who note that it is possible to be poor and not socially excluded or vice versa. Examples of the latter can be drawn from the Indian caste system, or else from classic cases of discriminated minority groups specialising in trade and commerce. Stewart et al. (2006, p. 5) elaborate on this, in tune with their focus on horizontal inequalities as a cause for conflict. They note that there are some groups who are privileged in some respects, yet still excluded from some important aspects of societal activity, particularly political participation, such as the Chinese in Southeast Asia or the Jews in Europe for many centuries. However, they make a distinction that those who are socially excluded are usually identified as having multiple deprivations, whereas these privileged groups suffer mainly from political exclusion, but they do not suffer multiple exclusions (ibid.). By making this distinction, they seem to reduce social exclusion (as opposed to ‘exclusion’ more generally) back into a metre of poverty, thereby discounting the valuable insight that derives from these examples, which is that exclusion and poverty do not always work together.

Similar observations have also been made in OECD countries. Atkinson (1998) makes the point that poverty does not necessarily always go together with social exclusion, and he argues that confusion of the two concepts is one reason for differences of view about the role of social security benefits (p. 9). Levitas (2006) also notes that, even in the UK, paid work itself may in some cases limit social ‘inclusion’ or that ‘economic inactivity’ does not necessarily lead to exclusion from social relations (pp. 123, 147). In such cases, when exclusion is associated with poverty, we are dealing with an overlap. However, it would be wrong to then integrate this overlap into the very definition of exclusion, thereby reducing social exclusion to its most restrictive case. That being said, once this distinction is recognised, there is still a tendency among these authors to reduce social exclusion back into some metre of poverty, similar to the case of Stewart et al. (2006) mentioned above.

In this regard, caution is also needed for the propensity to elaborate binary typologies of overlaps (e.g., poor and excluded; poor and not excluded; not poor and excluded; and not poor and not excluded), as this also succumbs to the tendency to treat exclusion as a state, similar to that of poverty, as discussed previously. This would also tend to compound the conceptual and methodological quagmires already associated with poverty and aggregated composite indicators. For instance, are we to develop a multidimensional ‘exclusion line’, distinct from a multidimensional poverty line? Many of these operational issues are the focus of contributions in Atkinson and Hills (1998), Burchardt et al. (1999), Hills et al. (2002) and Pantazis et al. (2006), among others. Notably, most of these authors tend to treat social exclusion as an outcome, which in turn leads us back to the slippery slope of ambiguous synonymy with poverty.

Causal and positional relativity

Indeed, this tendency to treat the identification of social exclusion in a similar manner as poverty leads to an arbitrariness that is far more intractable than the arbitrariness of threshold-setting in poverty measurement. Two often-overlooked dimensions of relativity help to shed light on this, which can be called causal and positional relativity. They are to be differentiated from the two meanings of relativity commonly referred to in the literature as mentioned in the last section, i.e., relative poverty (as per Sen) and contextual relativity (as per Silver).

Causal relativity refers to the observation that that there are multiple often contradictory processes at work within any condition or state of being, such that a person might face both exclusionary and inclusionary processes at the same time. In the IILS debate (Gore and Figueiredo 1997), this point was made by noting that processes of exclusion are not independent from processes of inclusion and patterns of integration; the ‘excluded’ are almost always ‘included’ in a variety of ways, such as Dalits in India who nonetheless serve important functions in the labour hierarchy, even if these are subordinated and exploited functions. Similarly, the inclusion of landless peasants into poorly paid and exploitative wage labour might be predicated on their exclusion from land assets. DuToit (2004) later contended that this is an issue of adverse incorporation rather than exclusion, which was then coined by Hickey and Du Toit (2007) as ‘adverse incorporation and social exclusion’. Across the linguistic divide of scholarship, however, Roulleau-Berger (1999) was making much subtler points in French, by describing how young minority people in France pass several times a day, and on a daily basis, through situations that could be described as either ‘excluded’ or ‘included’. Hence, the identification of exclusion in these situations depends on which causal process or situation, or even moment, we decide to focus on.

This is not simply another perspective on the multidimensionality of poverty. The conception of poverty as a state of being is more or less straightforward even if its measurement is not (i.e., lack of means or outcomes in relation to a relative or absolute metre, however measured). If exclusion is treated in the same way, thereby requiring a partial selection of criteria from which to determine whether someone is excluded or not, the risk is that the simultaneous and dialectical modes of integration and segregation that operate within social processes might be overlooked, particularly if these do not necessarily result in states that we might caricaturise as social exclusion in our abstractions. Even if we would wish to prioritise one process in order to characterise a person as ‘excluded’ in an absolute sense, such as ethnic segregation, how should we determine this priority over other contradictory processes? In some cases, the choice is made obvious by our normative concerns, although in many cases it is not.

Nor is this necessarily an issue of multidimensionality within exclusion itself. As noted in the beginning of this chapter, social exclusion is generally conceived as multidimensional, such as exclusion in various dimensions like health insurance but not from labour markets, etc. This point is also made in much of the scholarship and, indeed, some of the above examples above would fit into this mould, such as how exclusion from land assets can allow for exploitative inclusion in labour markets. Burchardt et al. (1999) also make a similar point with regard to their five dimensions of social exclusion in the UK (although the choice of these dimensions is debatable, as noted above). However, the multiple contradictory tensions could also be experienced within a single dimension, such as going in and out of situations of social integration or social isolation on a regular basis, as noted above in the studies by Rouleau-Berger (1999). As a result, the identification of exclusion possibly depends on the choice of process even within a single dimension.

The latter point is reinforced by positional relativity, meaning that exclusion is relative to a person’s position in a social hierarchy, both objectively and subjectively. In other words, a person can be an excludee and an excluder at the same time and will tend to perceive their exclusion (or inclusion) relative to the social strata immediately surrounding them. This is analogous to insights from the wellbeing literature that one’s subjective perception of wellbeing depends in part on one’s position or situation relative to one’s immediate surrounding, which serves as a basis of social comparison.⁷

This latter relativity can apply across many dimensions of comparison, such as ethnicity, class, caste, clan, occupations, gender, generation or location. For instance, a man might face forms of exclusion outside a household while at the same time practising forms of exclusion towards women within the household. Similarly, anti-Muslim activism by Tibetans in Western China presents a classic case of a subordinated minority group practising various forms of exclusion towards another subordinated minority group, partly as a reaction to their own perceptions of exclusion (see Fischer 2008a, 2009a, 2014a). Notably, to further complicate matters, Muslims are targeted precisely because of their economic success, and thus their own exclusion is difficult to corroborate with poverty.

In this sense, concepts of exclusion and inclusion reflect the constant jostling for social position within hierarchical social orderings. This conforms with the suggestion made by Room (1999, p. 172) that society can be seen as

a battleground of different social groups (based on social background, ethnicity, economic interest, gender, age, etc.), seeking to maintain and extend their power and influence, in a zero-sum struggle with other groups who they seek to exclude. ‘Exclusion’ is the result of this struggle, rather than a label to be attached to the casualties of some impersonal process of urban-industrial change. Social exclusion is a normal and integral part of the power dynamics of modern society.⁸

One (unidentified) participant the IILS debate summarised in Gore and Figueiredo (1997) made a very similar point, referring to the grey area between social exclusion and social inclusion, within which most people in the real world live most of their lives at one point or another.

In another formulation, which was classically Weberian, it was suggested that society could be seen as multi-layered, like a kind of staircase, with processes of exclusion and inclusion occurring at all levels and executed by various kinds of organized groups and associations who are seeking on the one hand to exclude others, to defend privilege and limit competition, and on the other hand to counter exclusion … It also implied that in seeing social exclusion as a cause of poverty, it was important to focus on the exclusionary and inclusionary processes occurring ‘at the bottom of the staircase. (Gore and Figueiredo 1997, p. 42–43)

This formulation is useful to think through how processes that we might identify as exclusionary can be constantly present for even those who live well above the poverty threshold, whether defined in relative or absolute terms, including among elites.

In other words, even a relatively secure person (from a poverty perspective) can psychologically experience insecurity in a manner similar to a poor person, even if, objectively speaking, they would be considered far from vulnerable.9 We treat the insecurity of a poor person with greater normative legitimacy, although from an analytical point of view we cannot assume that subjective emotive insecurity necessarily correlates with socio-economic insecurity. In some cases, there can even be an inverse relationship, when privilege induces grasping or implies having more to lose, whereas poor people might be more inclined to accept change (for a discussion of this in relation to rural subsistence economies, see Fischer 2008b, 2014a).

Both meanings of relativity throw a serious wrench into the definition or identification of exclusion, even once the norms of a society have been context-specified. However, they also clarify that exclusionary processes occur vertically across a social order, whereas poverty occurs horizontally at the lower end of a social order. In other words, all individuals or groups can perceive or experience an exclusionary process from whatever their position. Without belittling the normative importance of situations where exclusion overlaps with poverty, this clarification is important from an analytical perspective, that even the rich can experience exclusion. Hence, to focus on one aspect or process for the purpose of identification risks missing the simultaneous and dialectical modes of integration and segregation that operate within actions and events, the results of which may or may not lead to states of poverty – relative or absolute.

Dynamic divergences

A further observation that compounds the distinction between poverty and what we have come to call exclusion is that some exclusionary processes might in fact intensify with movements out of poverty. Migration in China serves as an intuitive example. The poorest in China today are typically characterised as rural residents whose livelihoods are based exclusively on agriculture and who possess some of the worst functionings or basic needs in China, such as in education, health or social security. In contrast, processes of exclusion – i.e., obstructed access to certain sectors of employment or social services (rather than a lack of employment or services) –are arguably faced most strongly during the migration of rural residents to urban areas through institutionalised systems of residency status (Ch. hukou). While these migrants might be relatively poor in urban areas (although in many cases they are not),¹⁰ in general they were relatively wealthy in rural areas before migration, in line with the widely accepted observation in migration studies that migrants, on average, tend to be wealthier, more educated and more entrepreneurial than the norm in their sources of emigration. In contrast, the rural poorest avoid these urban exclusions by virtue of remaining in agriculture. According to this logic, wealthier rural households – which tend to be more integrated into urban employment systems via one or more family members – would be more exposed to exclusion than poorer rural households. Similarly, movements out of poverty through the predominant vehicles of education and migration (both of which require considerable resources) might intensify rather than alleviate experiences of exclusion among rural households, which anti-poverty policies predicated on urbanisation would also tend to aggravate. These implications would be difficult to capture through standard income, basic needs, capability or relative measures of poverty, or even through inequality measures, given that they would occur at middle rather than lower social strata.

This point – that the dynamics of poverty and exclusion are often poorly correlated – is similar to criticisms of the ‘feminisation of poverty’ made by certain gender scholars, particularly Jackson (1996).¹¹ Her arguments hold a strong parallel with the treatment of social exclusion elaborated here, although elsewhere she also cautions against the integration of gender analysis with social exclusion (Jackson, 1999). However, this caution is based on her analysis of the way social exclusion has come to be conceived as a ‘binary and polarised formulation of inclusion and exclusion’ (ibid., p. 132). This caution is therefore very much in line with my own arguments regarding social exclusion.

Jackson (1996) provides a compelling argument for why gender is distinct from, and sometimes contradictory to, poverty and class. She contends that the ‘arguments which show how women’s subordination is not derived from poverty need to be excavated to demonstrate the (liberal) fallacy that poverty alleviation will lead to gender equity’ (p. 491). This is not to say that gender cannot overlap and reinforce poverty, as is variously emphasised by the scholarship on social reproduction or intersectionalities. Jackson acknowledges the importance of a gendered analysis of poverty, such as in the case of Indian households headed by widows that face higher-than-average levels of vulnerability and impoverishment (p. 493). However, she also qualifies that there is considerable evidence that gender relations are more equitable in poor Indian households than in wealthier households, in terms of women’s engagement with labour markets, contribution to total household income, control over income, and physical mobility. As a result, gender equity often appears to be inversely related to household income. She notes that studies of the Green Revolution have ‘shown a pattern of withdrawal of women’s labor from farm work and increasing dependence of women on men as household incomes rise’. Other studies have shown that ‘some of the most severe discrimination against the girl child in India is found in high-caste rural groups, characteristically also high income’; and her own research in Bihar found that ‘higher caste farmers had very few surviving daughters whilst the juvenile sex ratios in low caste and tribal households of the same village were much more balanced’ (p. 497).

She offers some explanations for these findings, based on a longitudinal study of a village of Uttar Pradesh in the context of the Green Revolution. These explanations place much emphasis on processes of structural change associated with development, such as employment changes due to crop changes and mechanisation, which reduced employment for women. As a result, rising prosperity led to the withdrawal or displacement of women from wage work, the strengthening of purdah norms and the inflation of dowries. The increasing dependence of women in upwardly mobile households therefore brought deeper aversion and higher mortality rates for girl children in these households (pp. 497–498).

Jackson clarifies that these findings are not to suggest that women are better off poor, but that this ‘is one way of looking at the limited degree to which poverty and gender development can be approached synergistically with the same policy instruments’ (p. 498). In other words, gender justice is not a poverty issue and this distinction is important to make given the tendency of development organisations to collapse all forms of disadvantage into poverty. She argues that rescuing gender from poverty analysis ‘involves poverty-independent gender analyses and policies which recognize that poverty policies are not necessarily appropriate to tackling gender issues because the subordination of women is not caused by poverty’ (p. 501). Moreover, she notes that non-poor women ‘experience subordination of different kinds … which make them important categories in their own right’ (ibid.).

Similarly, non-poor people experience exclusions of various kinds, which also make them important categories in their own right, without requiring an overlap with poverty in order to legitimate our attention. In particular, an exclusion that does not necessarily lead to any particular poverty outcome might still have a very powerful effect on various social processes of integration, even when it occurs at the middle or upper end of a social hierarchy.

In recognising this distinction, however, we must also beware of the contrasting danger of treating exclusion as synonymous with disadvantage, as warned by Room (1999, pp. 171–172). The challenge, then, is to differentiate social exclusion from disadvantage rather than from poverty. Similar to the circular logic associated with the capability approach, an exclusion may or may not arise from or lead to a disadvantage. Exclusion can create disadvantages (say, by excluding certain people from certain types of education, thereby leading to later disadvantages in labour markets), or else can be reinforced by existing disadvantages (such as when linguistic competency is used as a selection criterion in a multilingual setting dominated by one hegemonic linguistic culture, such as Chinese in China or English in the US). Distinctions between exclusion and discrimination could be made in a similar manner, although, in both cases, it brings us back to the question of why throw in the concept of social exclusion at all, rather than just dealing head-on with disadvantage and discrimination?

The benefits of differentiation

To give the benefit of the doubt, distinguishing social exclusion from poverty does provide some remediating potential to the concept of social exclusion by opening up potent avenues for the analysis of stratification, segregation and subordination, especially within contexts of high or rising inequality. These include at least three. The first is situations where exclusions lead to stratifying and potentially impoverishing trajectories without any obvious short-term poverty outcomes. The second is situations where exclusions among the non-poor shed light on obstacles to upward mobility faced by the poor. Finally, this distinction corrects the association between exclusion and conflict, as promoted scholars working on the closely related idea of ‘horizontal inequality’, and the tendency that derives from this to blame inequality-induced conflict on poor people.

Exclusions as processes of stratification and subordinated inclusion

The first avenue deals with understanding how processes of subordination, stratification and segregation can lead to various forms of disadvantage, discrimination or long-term poverty trajectories even when there are no obvious short-term poverty impacts. Such considerations are especially important in contexts of structural change such as urbanisation and migration, rising education levels or changing livelihoods patterns, during which the exact distributional implications of such changes might not be obvious, but where powerful stratifying social processes might nonetheless be at work. They also apply to situations where exclusions in certain domains allow for subordinated (or adverse) inclusion/integration in other domains, such as in labour markets or in cases of financial exclusion/inclusion, as noted earlier. This latter point is similar to the concept of ‘adverse incorporation and social exclusion’ proposed by Hickey and Du Toit (2007), except that their treatment is restricted to the space of poverty. The approach proposed here widens consideration to processes occurring vertically across social hierarchies.

This strength particularly applies in contexts of rising inequality or polarisation. In terms of income distribution, rising inequality effectively results in a flattening-out of the income distribution, with fewer people in the middle of the distribution and more at the tail ends (mostly the bottom tail end). In the process, middle strata potentially face the greatest relative downward displacements. Lower strata will experience more competitive pressures due to the downward displacements of those from above, or reduced opportunities for upward mobility, and this will induce considerable churning within the lower strata.¹² However, as discussed previously, many poor might be insulated from churning at higher levels, such as rural dwellers with little or no integration into urban labour markets. Among those in the lower strata who do experience exclusion, they will more likely experience it as obstruction of upward mobility, while their relative position within a social hierarchy might remain unchanged. Rather, the greatest insecurity in terms of loss of relative position (rather than poverty) is usually faced by various middle strata. It is in this sense that we often see intensifying exclusionary pressures among middle strata in contexts of rising inequality.

An example of this comes out of my own work on Tibetan areas in Western China, in my attempt to understand intensifying exclusionary processes within a context of rapid but dis-equalising growth, falling poverty, rising education levels and other developmental improvements (see Fischer 2009a, 2014a). Most conventional measures of exclusion offer little insight into this situation, except perhaps inequality-based measures given rapidly rising inequalities alongside rapid growth. However, inequality measures put the focus on the poorest strata of Tibetan society, whereas I came to realise during my fieldwork that some of the most intense exclusionary pressures – as well as grievances and political frustrations – were faced by relatively elite and/or upwardly mobile Tibetans, such as Tibetan high school and university graduates (only about five per cent of the population had a high school or university level of education at the time, in the early to mid-2000s). In particular, the implementation of competitive labour market reforms and educational campaigns exacerbated exclusionary pressures among this elite educational stratum by accentuating the linguistic and cultural disadvantages faced by these graduates in competing for public employment correspondent with their educational achievements and employment expectations. Notably, these particular pressures were not faced by Tibetans with lower levels of education. The ‘excludees’ in this case had among the highest educational achievements of their respective communities, they came from families with the resources to be able to finance these levels of education and they had the ability to perform relatively well at these levels (although not enough to compete with Chinese graduates). As discussed below, the resultant exclusions offer important insights into recent tensions in this region, in addition to the more blatant proximate causes such as discrimination or political repression.

Indeed, it was these observations that led me to rethink the social exclusion approach, precisely because the concept of exclusion seemed salient to describe experiences on the ground, even though these experiences were not reflected through any of the conventional absolute or relative poverty measures. Moreover, while this example definitely implicates practices of discrimination, it also brings to light how processes of structural and institutional disjunctures can lead to effective discrimination, even though discrimination might not be necessarily intentional. The methodological challenge, then, is to find ways of identifying these structural and institutional disjunctures across hierarchies and to differentiate them from intentional practices of discrimination. Notably, similar challenges also confront work on identifying social and economic rights abuses or on structural violence more generally.

Another poignant example of this application is in the study of immigration. For instance, when more stringent rules and procedures are imposed on immigrants to North America or Europe, including profiling or the criminalisation of illegal immigration, it is unlikely that the increased stringency stops – or is even intended to stop – such immigration. Immigrants continue to be demanded in a widening variety of sectors of employment, from agriculture to services (exemplified, for instance, but President Trump’s use of migrant labour despite his anti-immigrant rhetoric). Hence, the tightening rules are unlikely to exclude immigrants, in an absolute sense. However, stringency allows for stronger mechanisms of subordination and segregation during the integration of such immigrants into the receiving labour hierarchies. This result might or might not be the intended purpose of such rules – the rules themselves might have evolved out of impulsive political reactions within the receiving societies, themselves undergoing a variety of exclusionary displacements among middle strata due to rising inequality. The important analytical point of this example is that these processes often have little correlation with poverty, given that the targeted immigrants are often well-educated and are often not poor, even according to the standards of the recipient countries, particularly if they find work. Nonetheless, despite the lack of correlation with various measures of poverty, these exclusionary processes are very important for understanding the resulting stratification of labour hierarchies, which could well lead to future trajectories of impoverishment, discrimination or disadvantage.

Obstacles to mobility

The second related strength of distinguishing exclusion from poverty is that the obstacles faced by poor people attempting to escape poverty through upward mobility are clarified by exclusions occurring in the non-poor social strata that these poor are attempting to enter. This is especially important in situations where poverty reduction strategies are predicated on upward mobility, such as education or entrepreneurship, versus improving the terms of labour. For instance, the idea that education is good for poverty reduction is largely based on the presumption that those receiving education will subsequently move into higher strata of employment, such as from farming, menial wage labour and informal petty trade, to formal white-collar employment, whether in the public or private sector. However, this idea is problematic when these targeted sectors of employment are already subject to strong exclusionary pressures. Indeed, this insight puts into question the mainstream human development emphasis on absolute levels of education without corresponding emphasis on employment generation and upgrading, particularly in contexts of high or rising inequality, as noted in the first point above.

This perspective is different from the more common assertion that we need to understand the relationship of more advantaged groups to the socially excluded, as suggested by Room (1999, p. 172), which is still based on a binary conceptualisation of excluders and excludees. Rather, here the emphasis is on how exclusions experienced by those higher up in a social hierarchy can lead to a variety of knock-on effects lower down in the social hierarchy. Warren and Tyagi (2004) make this point with regard to the importance of looking at the pressures on the middle classes in the US in terms of how this affects poor people aspiring to that middle-class status. It is also evidenced in my Tibet research mentioned above, whereby the difficulties experienced by Tibetan and other minority graduates in obtaining appropriate formal employment were in part due to labour market reforms occurring more generally across China, which intensified competition within such employment between more advantaged Chinese graduates. Related points have also been made with regard to gender by Jackson (1996, p. 501), who argues that the experiences of non-poor women are relevant to poor women through role-modelling and changing social norms, in both positive and negative ways. Similarly, exclusions occurring within middle strata could have ideational and demonstrative influences on the poor, such as by signalling the importance of cultural assimilation versus political assertions of language rights and affirmative action.

Inequality-induced conflict

The third avenue is in the study of conflict, particularly with respect to correcting the common tendency in much of the literature to implicitly blame inequality-induced conflict on poor people. For instance, if increasing inequality is evoked as a cause of conflict without any further qualification, the implied presumption is that rising inequality raises the relative disparity of the poorer sections of the society in question, thus raising their discontent and their propensity for engaging in conflict.

Perhaps the best example of this is found in the ‘horizontal inequality’ approach, referring to inequalities between groups rather than between individuals, as laid out by Stewart (2002). More specifically, Stewart et al. (2006) make an explicit effort to connect social exclusion to their broader project of identifying ‘horizontal inequalities’ as a critical determinant of inter-group conflict. However, they operationalise social exclusion as involving multiple overlapping deprivations and argue that, because of economic deprivation, the ‘[socially excluded] appear to have little to lose by taking violent action’ (Stewart et al. 2006, p. 6). Interestingly, the only solid examples used by the authors to substantiate this contention are cases of discriminated, typically minority, cultural or religious groups, including, among others, Tibetans. Here again, we must question whether those who do agitate are necessarily suffering from severe deprivations due to multiple exclusions. For instance, as discussed above, the Tibetan case actually suggests that the aggrieved socially excluded are, in many cases, people who are best characterised as middle strata, squeezed by both downward competitive pressures and obstructions to upward mobility.

Moreover, we know from actual studies of conflict that most conflicts involve a considerable degree of elite participation, especially in leadership and also in core support. Obviously, there is no doubt that the poor often serve as a reserve army, literally or financially. But the very poor are often too poor or too unhealthy to engage in sustained violent conflicts. Similar to migration, such engagement usually requires resources and organizational capacity, things that the very poor presumably lack.

Indeed, Stewart et al. (2006) qualify their argument in this respect, noting that leadership plays a critical role in emphasising and accentuating particular identities, and that leadership emerges out of the middle classes rather than the deprived in most of the conflicts they studied (ibid., pp. 7–8). However, the problem is that they end up treating different strata in arbitrarily different ways with respect to how common grievances might coalesce within a common cross-strata group cause. Lower strata tend to be treated in a functionalist or even primordial manner with respect to grievances or cultural affinities, while leadership is treated in an instrumentalist manner, using these grievances and affinities to enact political ends (e.g., see pp. 9–10). Perhaps as a means to overcome this dualism, the authors refer to a complicated mixture of exclusions, although in their attempts to operationalise social exclusion as poverty, they appear to convolute rather than enhance their otherwise-interesting discussion of conflict.

Rather, by focusing our attention to include exclusionary processes occurring in the middle and at the upper ends of a social hierarchy, we can understand why the non-poor might also come to be aggrieved by rising inequality, thereby clarifying inequality-induced social dynamics that might contribute to conflict. In particular, as discussed above, the dislocating effects of rising inequality can be seen to be most intense at the middle strata of a social hierarchy in terms of relative downward displacements. In the event that these displacements are caused by exclusion, whether perceived or real, these experiences can turn into potent focal points for grievance and rallying points for political agitation, particularly when they occur among politically active people. This perspective is important because it implicates middle classes and elites into an understanding of how inequality might induce conflict. Indeed, it resonates with the theses of Polanyi (1944) and Arendt (1951) regarding the social origins of fascism in Germany – they both identified economic insecurity among the middle classes as a critical factor. It is precisely this angle that gives a concept of exclusion that is differentiated from poverty an edge over poverty and inequality approaches in understanding social conflict, given that the methodologies of poverty and inequality analysis tend to divert attention away from vertically occurring processes that cut across hierarchical social orderings and, in particular, away from elites.

Instead, by conceiving social exclusion as vertically occurring processes of obstruction or repulsion, we come closer to prying open the puzzle of how social dislocations might occur in very different ways across different strata of an intra-group hierarchy, yet still might provide for a basis of common grievances within the group. For instance, this approach addresses the puzzle posed by Mann (2005, p. 5), regarding why, in situations of murderous ethnic cleansing, class-like sentiments come to be channelled into ethno-nationalist ones. This requires an understanding of the possibility for vertical social bonds to form in the face of a commonly perceived adversity, crossing over the horizontal loyalties that are presumed to exist in conceptualisations of class and other forms of socio-economic status. The actual experience of exclusion or insecurity would differ considerably across these horizontal strata, but processes of exclusion that cause deprivation at the lower end of a social order can, in many cases, also cause an erosion of economic or political power or social status among certain elites.

Hence, a heightened sense of insecurity can occur among upper social strata in line with multiple deprivations among lower social strata, thereby allowing for a perception of shared adversity that forges vertical bonds of solidarity within a group, in contrast to differences in actual experiences faced by various social strata within the group. Common perceptions of cause, rather than common outcomes, allow for common narratives to emerge across social strata, which can then be used as mobilising focal points for organising common remedial strategies. This can be observed, for instance, with the rise of populist right-wing parties around the world, which tap into classic working-class concerns while also finding support and leadership among various elite factions (as did fascism in the 1930s, for that matter). This understanding of exclusionary processes therefore helps to move towards a more nuanced position that understands the actions and ideologies of both rich and poor alike as both instrumental and normative at the same time.

Conclusion

This chapter has sought to highlight the conceptual ambiguities of the social exclusion approach through a deconstruction of the concept. Despite the fact that most of the literature agrees that the value-added of social exclusion is found in its treatment of processes and that social exclusion can occur without poverty, most attempts to operationalise the concept end out reducing it to a description of certain aspects of poverty. It is usually formulated in terms of multiple and cumulative disadvantages that lead to or reinforce multiple deprivations, or else as a description of various social aspects of deprivation. This opens the way for valid criticisms that the concept is redundant with respect to already-existing concepts of poverty, particularly more multidimensional and processual concepts of poverty such as relative or capability deprivation.

However, similar to criticisms of the feminisation of poverty by certain gender scholars, social exclusion must be differentiated from poverty in order to do justice to the study of exclusion (or what we call exclusion), given that exclusion and poverty do not always go together. This insight is important because an exclusion that does not necessarily lead to poverty might still have a very powerful effect on social processes such as stratification, segregation, subordination, discrimination or conflict. Indeed, exclusionary processes at the upper end of a social hierarchy are especially powerful. Such processes are therefore pressing concerns in their own right and should not require an overlap with poverty in order to legitimise our attention.

The remediating benefits of making this distinction are on the insights that it can provide on many processes of subordination, stratification and segregation, particularly within contexts of high or rising inequality, which are not effectively captured by poverty or even inequality methods of analysis. In other words, absolute and even relative indicators often tell us little about processes of exclusion and marginalisation. If they do, they usually only do so by providing clues about the spaces within which exclusionary processes might be operating. Indeed, standard statistical sampling methods based on outcome indicators might be poorly suited for capturing the processual emphasis of the social exclusion approach, which would be better served by more inductive methods that are able to trace subtle social dynamics rather than cross-correlations across divergent people. This would include interdisciplinary analyses of structural and institutional disjunctures and asymmetries operating not only across social hierarchies, but also among comparable cohorts within a social hierarchy, for instance, with similar levels of educational achievements and employment expectations. Ultimately, the social exclusion approach also calls for a shift of methodological and even epistemological dimensions.

Most importantly, exclusion understood in this way avoids the tendency to blame poor people for a variety of perverse social dynamics that emerge across social hierarchies in response to rising inequality, which can be easily misattributed as stemming from poverty due precisely to their association with inequality. Conflict is an obvious example given the common assertion in academic, policy and journalistic circles that increasing inequality will exacerbate conflict. While this might be true, it is equally important – indeed, it is an ethical responsibility – to also remind ourselves that the conflicts to which we refer usually involve substantial elite participation, particularly in leadership positions. Hence, we need social theory to address how elites themselves might find grievance with inequality, lest we fall into a trap of crude instrumentalism. The poor have enough to deal with; they do not need the additional burden of our implicit blame or paranoia.

The question still remains whether all of these functions of the social exclusion approach could be even more effectively analysed with existing sociological concepts, such as many of those mentioned above. The addition of the concept of ‘social exclusion’ serves no obvious additional purpose and brings no obvious value added to these concepts. Indeed, whereas we can resolve the issue of whether social exclusion is redundant with poverty, we are then left with the question of whether it is redundant with concepts of disadvantage or discrimination. Moreover, the approach has brought a lot of vagueness and confusion, and even though it initially seemed to offer some exciting potential for bringing in more critical perspectives into the field of poverty studies, it has also exhibited a strong propensity to be coopted into various mainstream agendas around inclusion. Since most of the debates of the 1990s and early 2000s were wrapped up, its application with respect to developing countries has been mostly through these agendas.

The challenge to rehabilitate the approach is to emphasise the potential for a critical political economy element within it, such as through an understanding of exclusion as part of the exercise of power within social relations, or that exclusion in certain spheres or dimensions creates dependence in others, which becomes a basis for exercising power. In this way, it permits a discrete consideration of social power relations and the ways these work to confer or deny access to resources or opportunities. Nonetheless, the danger in even this emphasis is in replacing existing, established and relatively clear social-scientific concepts and methods with vague and unspecified new ones, particularly when these new ones are so easily coopted away from their original intent.