Understanding Poverty, Defining Interventions: Why Social Relations Need More Attention in Livelihoods Analyses and Why This Complicates Development Practice

In: Livelihoods and Development
Author: Urs Geiser
Full Access

Linking an analytical understanding of why people suffer from inequalities with the sphere of policy trying to fight such inequalities is a core ingredient of the livelihoods approach. However, its application is increasingly criticised for being apolitical and inspired by neo-liberalism. The present chapter argues, though, that these critics risk throwing out the baby with the bath water. Its core focus is to propose that the challenge is not the absence of attention to power or politics in livelihood-centred studies, but how power and politics are conceptualised in these studies, and how power and politics are then addressed in the sphere of interventions. Using Bernstein’s ideal-type juxtaposition of residual and relational ontologies in development studies, it shows that both, indeed, do address issues of ‘power to’, ‘power with’, and ‘power over’ – though in distinct ways. But mainstreamed readings of the residual (i.e. neo-liberal) and the relational ontologies (i.e. orthodox Marxist) – though both claiming to be political – equally fail to unravel the everyday production of inequality. Empirical insights from rural Sindh in southern Pakistan illustrate that both these takes operate with pre-conceived typologies of the rural social universe (the ‘small farmers’, the ‘state’, ‘the peasants’) and that both neglect a differentiated engagement with the actual practices of social relations as experienced by marginalised people in their everyday life. Therefore, the chapter calls for a re-energised critical relational livelihoods perspective that gives due attention to agency and structure, to a more differentiated and site-specific stratification of the rural populace as well as the heterogeneity of ‘the state’, and a context-specific understanding of inequality as an outcome of multifaceted social power relations between the multitude of social groups and the heterogeneous state system. Such an analysis also calls for more differentiated policies and practices to challenge inequalities, beyond the development industry’s mainstream.

Introduction: The Argument

Livelihood approaches provide an essential lens on questions of rural development, poverty and well-being, but they need to be situated in a better understanding of the political economy of agrarian change.

scoones 2015: xvii
This paper engages with contemporary uses of the ‘livelihoods approach’ in the broad field of rural development. Rural development remains a core engagement of many researchers and practitioners who are deeply concerned about persistent poverty and inequality in many parts of the world. They are interested in the everyday challenges facing people, and therefore attempt to better understand and analyse the reasons so many still have to struggle to make ends meet. It is these in-depth insights – gained through a livelihoods approach – that should aid the design of practical and effective interventions to be implemented by development practitioners ranging from state agencies and non-governmental organisations to supporting donors. A livelihoods approach focuses on people so as to learn how they construct their livelihoods using the assets they control. It understands people as active agents, but it also addresses the wider social, cultural and political context in which they are embedded – a context that may support some people’s livelihood-related efforts while hindering others. These insights are then intended to help organisations reflect on how best to support people’s own efforts. Thus, the linking of analytical insights with the sphere of policy is a core ingredient of the livelihoods approach.

Analysing the causes of poverty and defining interventions are intrinsically interwoven. The issues uncovered by research should, in principle, be addressed in practice. Issues which that analysis does not highlight will usually feature less in the process of defining interventions. This statement may, of course, be oversimplistic and a reification of a linear and positivist understanding of the science–policy nexus (Keely and Scoones 2000; Leach et al. 2007: 5), but it does hint at the need to at least reflect on the relationship between understanding and action. After all, the notion of making development more ‘evidence-based’ has become an important mantra within the development industry.

This interest in the relationship between understanding and action, evidence and intervention, forms the starting point for this chapter’s engagement with the livelihoods approach. As a matter of fact, livelihood research has become an integral part of contemporary attempts to understand why some people are in a position to maintain or even improve their living standards while others not. And insights from this research are influential – at least to some extent – in shaping development policies and practices. This kind of development policy, based on evidence and insights from livelihoods research, is no longer limited to progressive non-governmental organisations, but has been mainstreamed within the bilateral and multilateral aid community, for example, the Food and Agriculture Organisation’s (fao) programme to ‘increase the resilience of livelihoods to threats and crises’ (fao 2016) and also the Livelihood Asia Summit 2015 (access 2015).

However, it has also come in for criticism. Development research and policy inspired by a livelihoods approach is increasingly criticised for being apolitical, that is, for not sufficiently addressing the power relations that keep people in poverty in the first place. Such apolitical analyses are accused of fostering an apolitical agenda in development intervention. This criticism will be outlined further below, and it does indeed highlight trends in how livelihoods frameworks are currently applied in certain sections of academia and among development practitioners. However, I argue that this criticism does not disqualify the livelihoods approach per se. The core focus of this paper is to propose that the challenge is not the absence of attention to power or politics in livelihood-centred studies, but how power and politics are conceptualised in these studies, and how power and politics are then addressed in the sphere of interventions. I argue that most researchers and practitioners do include issues of power and politics in their endeavours, but they imbue these terms with different meanings that are in turn embedded in different ontologies. The issue, then, is not to disqualify one or the other as apolitical, but to examine, first, their respective strengths or limitations in explaining the importance power and politics have in restricting or enabling people’s livelihood options (in specific contexts); and second, how these insights can inform debates on interventions in a meaningful way.

In order to characterise such differences in ontologies I build upon Bernstein’s famous typology of residual versus relational understandings of rural challenges. This typology nicely positions two fundamentally different bundles of analyses and action. They have very different – almost dichotomised – takes on power and politics, and each inspires a clear-cut template for how to support poor people’s livelihoods. Both talk about social relations and are thus addressing power and politics, but I hope to show that their way of categorising the rural universe, and their conceptualisation of social relations among these categories often favour ideological assumptions over the nitty-gritty of lived experience.

Beyond this rather dichotomised debate, though, many concerned researchers and practitioners are engaging with more differentiated conceptualisations of the political-administrative, economic and socio-cultural fields within which people have to construct their livelihoods. They base their research on ontologies that allow for far more differentiated readings of these contexts and their relations to people’s life-worlds. As a result, they highlight the complexities and context-dependency of social relations. However, unlike in the mainstream positions mentioned above, analytical insights from these differentiated analyses have no clear-cut templates for informing practice. Their insights are complex and not easy to translate into practical interventions, not least because those assumed to be in charge of policy and policy implementation – the state, non-governmental organisations (ngos) and donors – are not excluded from the analyses of social relations. On the contrary, the role they play in the production and reproduction of inequality is a focus of particular attention.

To give a simple summary of my argument: first, accusing the livelihoods approach of being apolitical is a distraction. Rather than asking whether or not power and politics are addressed, I consider it more useful to look at how power and politics are conceptualised and how power and politics then become a theme of interventions. Second, the more differentiated power, politics and social relations are analysed, the more challenging it becomes to define interventions for ‘development’.

In order to illustrate and, above all, substantiate my argument, I start in the following section with the case of a study report I came across during recent fieldwork on rural labour in the Sindh province of Pakistan. In the next section, I shall draw on my ongoing research in the same region to show that this report does indeed address dimensions of power and politics, but in a specific way. It occludes certain relational issues of power, particularly the relations between different categories of farmers, between farmers and sharecroppers, and between all of these and the state. In the subsequent section, I will address the question of whether this makes this report apolitical; to do this I shall briefly discuss what ‘apolitical’ means in contemporary critiques. Then, in order to substantiate my claim that this criticism misses important points, I shall present Bernstein’s residual versus relational understandings of rural development in the next section. In two subsequent sections, I will show that the debate on the apolitical nature of the livelihoods approach can be located between these dichotomised positions. Thereafter, I turn to endeavours that go beyond the above dichotomy and study the analysis–action nexus with a greater degree of differentiation. Once again, these examples refer to farming and sharecropping in Sindh. They also underline my argument that translating differentiated insights into social relations complicates the task of defining interventions. The final section then argues for a critical and relational livelihoods perspective that does not shy away from complicating, and even challenging, the sphere of interventions.

The Case: Analysing Vulnerabilities in Rural Sindh, Pakistan

The issue of vulnerable populations is a problematic one … People’s susceptibility to suffer damage is embedded in everyday power relations and political economy and can be inflected by class … gender … and ethnicity … among other factors.

Social Policy and Development Centre [spdc] 2015: 2

The quantum of income, or poverty level are not very good indicators of vulnerability. What matters more is the stability and diversity of income.

spdc 2015: 48

The first quote introduces the general analytical direction, the second highlights one of the findings of a study in the Sindh province of Pakistan, the final report for which was published in early 2015. The study was undertaken by a Pakistani think tank and financed by well-known international donors. Its objective was to understand the vulnerability situation of rural households with an emphasis on women and their (lack of) capacities to cope with climate change. Data were collected through a large sample of household interviews located in four different agro-ecological/livelihood zones, i.e. two in regions with irrigated agriculture, one in a rain-fed agricultural zone, and one on the coast where the main livelihood is fishing. The data collected encompassed various types of assets including education, incomes, access to social networks, infrastructure and employment, and gave specific attention to the interviewees’ own perception of vulnerability and coping capacities. Data were comparatively analysed first at household level, then at community level, and finally at the level of the agro-ecological/livelihood zone.

The findings indicate a close correlation between vulnerability and (lack of) access to transportation, education and health facilities, the widespread nature of agricultural wage labour and the limited alternative sources of livelihoods. It also refers to questions of access to land and water:

Agriculture landholding at family level has become skewed due to distribution of land among multiple legal heirs: this makes the size of landholdings too small for productivity and profitability. Water scarcity is also a major cause of reduced farm activities leading to high unemployment rate at village level.

spdc 2015: 44
In sum, the study found that

access to services (health, education) and roads was positively associated with vulnerability, education partially so, but quantum of income or poverty level was not a good indicator of vulnerability – more significant was the stability and diversity of income.

spdc 2015: xiv, see also 91–92
As a consequence, the report’s recommendations call primarily upon the state and ngos to invest in physical and educational infrastructure in order to enable more diversified income opportunities. It ends by formulating the core task ahead as the need ‘to support the ingenuity and resourcefulness, of both men and women to negotiate the challenges they are facing and will face’ (spdc 2015: 94). Regarding the study’s analytical approach, the authors conclude by reiterating the ‘significance of nuance, context and attention to class and gender differentiated analysis’ (spdc 2015: 94).

Though it was not explicitly stated, the study followed a contemporary livelihoods approach. It started with an analysis of individuals’ lived experiences at household, community, and regional levels. It researched their assets (including social capital), their access to assets, and contextualised these insights within wider development aspects such as the state’s institutional capacities. The study also addressed issues of vulnerability and coping strategies – again core notions in livelihood-oriented research. The findings reflect contemporary concerns as well: a concern for intra-household inequalities, especially gender; the limitations of the physical resource base, for example water and soil quality; the importance of livelihood diversity (the ‘watchword’ of the livelihoods approach according to Scoones [2009: 2]); the role of social networks; and the importance of the state. Finally, the study addressed ‘everyday power relations and political economy [which] can be inflected by class … gender … and ethnicity … among other factors’ (spdc 2015: 2). But though the study looks at power relations and their impact on livelihoods, I argue that it misses out important facets of power relations and thus other impacts on livelihoods.

The Complexity of Social Relations

The report I have briefly summarised above serves as an inspiration to reflect more generally on the relationship between analysis and action. After all, the report is based on a thorough analysis of everyday livelihood challenges faced by people in specific regions of Sindh. Based on these insights, it makes clear recommendations to the state and to ngos for interventions. Thus, it provides relevant and novel insights into important facets of the challenges different categories of people at the grassroots have to face. However, it misses important dimensions of the relations between these different categories of people. The report explicitly takes note of power relations, for example in the sphere of gender, but other power relations faced by the rural population in Pakistan’s Sindh province are not addressed.

To substantiate this critique, a brief discussion of these other facets of political economic realities in Sindh is due, based on my ongoing research into struggles for land reform in the region. This research, in turn, follows up earlier research into the social relations of production around natural resources (e.g. Shahbaz, Geiser, and Suleri 2013). Here, I concentrate on an irrigation-based agro-ecological/livelihood zone analysed in the report under examination.

My discussion specifically takes into account the skewed distribution of land holdings and the close relation of this inequality to landlordism on the one hand and frequently exploitative forms of sharecropping on the other. Though it emphasises agricultural production by farmers, the study summarised above does not differentiate between farmers with different amounts of land. Though addressing issues of social networks, it did not ask for the relation between landowners and sharecroppers or wage labour. The study does address class, but it understands class in a more static way, using proxies such as categories of income or gender.

The first critique concerns the notion of ‘farmers’, and especially ‘small farmers’. Data published by the official agricultural census (Bureau of Statistics 2011: 82) hint at the reported distribution of land, which indeed suggests the widespread presence of small farmers. But it hides the true extent of unequal land distribution. This discrepancy is due to the fact that for the purpose of official census enumeration, landowners often divide their land among their large capital of family members and relatives, making landholding sizes appear to be smaller. Consequently, a landlord is disentangled into a bundle of ‘small farmers’ (see also Shaikh [2001: 131] on the use of this coping strategy in the context of earlier land reform efforts in Pakistan).

My own interviews, and in particular studies undertaken by concerned local organisations and researchers (e.g. Shaikh 2001; Budhani and Gazdar 2011), show that there is indeed a category of what one could label small farmers, that is, people that own and operate up to around say 5 acres of land and are just able to eke out a living with a low level of productivity. Many of them, though, have to supplement their land-based income with casual labouring whenever they can, including working on their fellow farmers’ land. A large category owns between 10 acres and roughly 50 acres. While some of them cultivate part of this land as owner-operators, most already employ one or more families of haris, i.e. sharecroppers. Then there are those with ‘much land’ – as people say – in reference to landlords (often called zamindars) who control hundreds and thousands of acres. The notion of control is used here because actual land ownership is difficult to assess. These landlords, whose lands are toiled by dozens of hari families, are often absent, and managers (kamdaars) look after their farms.

This brings me to the second critique, that is the notion of agricultural wage labour mentioned in the Sindh study. Haris are a phenomenon that is quite specific to Pakistan’s Sindh province, but the word is not mentioned once in the report. Hari families are engaged by larger landowners to cultivate land on a sharecropping basis. Thus they are not, strictly speaking, wage labourers. While in earlier times the landowner would supply land, water and seed as inputs, and the haris labour and oxen, these arrangements have become highly differentiated and complex in the context of Green Revolution–induced changes to cultivation practices. The harvest is shared (a moment called batai) between landowner and haris. While sharing used generally to be on a 50–50 basis, this too has become complex and now depends on the (informal) share arrangements around the range of inputs required for modern-day farming. Another thing that is worth mentioning is that, after harvesting, the produce is usually marketed by the landowner or farm manager, but the haris rarely know the price obtained.

The third (and main) critique refers to the study’s lack of attention to social relations. Though de jure a Tenancy Act was introduced in 1950 to protect the rights of sharecroppers vis-à-vis landowners (Government of Sindh 1950; Hasnain and Marri 2015), de facto sharecropping arrangements are not formalised, and thus the Act cannot apply. As a matter of fact, during my interviews, I did not find one hari or small farmer/labourer who knew about this state legislation. Haris work at the will of their landlords, often without any written contract. They are allowed to build a temporary house – or, to be more precise, a hut – somewhere on the fringes of the land, but have to vacate the place once the (informal) sharecropping agreement comes to an end. Some of the hari families stay with a landowner for a generation and more, while others have to move rather frequently. But above all, evidence collected by several concerned organisations hints at the enormous degree of dependency (e.g. Hussein et al. 2004; Human Rights Commission of Pakistan [hrcp] 2015). While in some cases and in some regions of Sindh, relations are still guided by a ‘moral economy’ based on patron–client relations (Hussain and Mohyuddin 2014), in many cases haris are completely at the mercy of their lords or the kamdaars. This dependency usually stems from deep indebtedness. Today’s agricultural production needs credit, and credit is required for the reproduction of labour too, for example, medical emergencies, food during off-season and marriages. Landlords advance credit to the haris, but without written records. As mentioned above, haris also have no control over, or information about, the sale of the produce. As a consequence, they are often unable to come out of debt. This leads to the phenomena of bonded labour, and several highly committed local organisations have to engage with the worst forms of physically punished bonded labour (e.g. Green Rural Development Organisation [grdo], n.d.).

The fourth critique is linked to social relations, and concerns access to land. Here, the report discussed in the second section does not mention the objective reality of haris, but refers to the inheritance challenges faced by small farmers. It highlights the fact that inheritance reduces land sizes and implies the need to search additional labour opportunities and diversify income sources. Of course, issues of inheritance, the related subdivision of land holdings, and the lack of labour opportunities are important factors. But they mainly affect the aforementioned category of smaller landowners/owner-operators, and the others much less. But above all, land inheritance is not an issue at all for the masses of haris, as there is no land to be inherited, not even the plot where the hut is built.

A fifth and final critique concerns access to water. Access to water is indeed a challenge, but not for everyone. Huge irrigation infrastructure brings water from the Indus River to the fields of farmers, both smallholders and larger landlords. As in many of these state-controlled schemes, water distribution is highly politicised. Almost as a rule, tailenders suffer, while larger landowners are often found closer to the main canals. They also tap the main canals directly (and illegally; see also Indus Consortium 2015: 23). But the problem goes further. To be entitled to a share of water, a landowner needs to be a member of the Water Users Association along the canal he/she depends upon. To become a member, one needs the land title certificate issued by the state’s Board of Revenue. However, many smaller farmers or political opponents do not have such a certificate or cannot get one. Last but not least, many (though not all!) among the government’s irrigation staff or revenue officials are known to react positively to specific incentives (see also Indus Consortium 2015: 24–27).

This concludes my critique vis-à-vis the analysis of rural livelihood challenges as portrayed in the (otherwise highly relevant) report under discussion. I strongly argue that these political economic issues around relations between different categories of social groups or classes, and between these and the various agencies of the state, structure the everyday life of rural households, providing opportunities, but in most cases producing and reproducing constraints and hindrances to people’s ability to achieve sustainable livelihoods. They also create vulnerabilities to climate change.

In the following paragraphs, I turn to the recommendations regarding interventions in the report under discussion, based on the research insights it laid out. In sum, the report concludes that there is a lack of stability and diversity of income, as is clear from the quote at the start of the second section. The key challenges are seen as being in the spheres of assets and access to assets, that is, access to health and education facilities, physical infrastructure (for example roads), agricultural extension and markets. These are the challenges facing local communities of small farmers, and within these categories, women in particular. Therefore, the report recommends improving access to health and education facilities for lower income groups and women, and road infrastructure. These improvements are to be addressed through state and ngo interventions. Thus, just as the problem analysis focuses on issues of access to assets by small farmers, its recommendations focus on providing assets and access to these assets. Of course, providing better assets to small farmer households is important though not easy, but it is, I would argue, the main thrust of most rural development interventions of the last decades, and thus much experience is available, in theory.

My point here is that issues of access to assets are important, but the larger question is ‘for whom?’. I argue that a more differentiated attention to the classification of social groups and to the social relations between these groups (haris, landlords, kamdaars, the field staff of the Board of Revenue and the Irrigation Department) would lead to a rather different set of recommendations for interventions (I will show these in one of the last sections). As illustrated above, the rural social universe in many regions of Sindh is characterised by considerable tensions. First, it shows that the category of small farmers exists, but that it needs to be understood in relation to other groups, above all the landlords and haris. The vulnerability of haris is not primarily dependent on physical infrastructure or education, but caused mainly by their social relations with their landlords. Even if education or infrastructure were improved, in all likelihood this would not change the suffering of haris at all. It would almost certainly be particular categories of landowners (probably those owning over 10 acres), their land managers and their networks of relatives who would stand to benefit. Social relations also encompass the state and ngos. The report under examination takes these actors to a great extent as external to the rural life-world, and thus they are to be entrusted with providing support to the rural poor. However, the state is deeply entrapped in the reproduction of poverty, as follows for example from the problems described above of access to water by some sections of rural people and not others, the lack of access to land titles by some and not others, or the inefficiency of the Tenancy Act. Again, these may not be new insights, but the report under discussion reiterates the need to continuously insist on the importance of social relations in rural development (see also Harriss 2007 and Mosse 2010). By way of a reminder, Lieten and Breman (2002) have already critically discussed planned interventions by the Asian Development Bank (adb) in rural Sindh. The adb’s plans were to address rural poverty mainly through the provision of physical infrastructure: very little attention was paid to haris, and no interventions were planned to address the root causes embedded in social relations.

The next section engages with the question of whether the report by spdc (2015) illustrates an ‘apolitical’ application of the livelihoods approach or not.

The Livelihoods Approach and Power Relations: Throwing Out the Baby with the Bath Water

As mentioned in the introduction, some would argue that the report I discussed is a typical example of an apoliticised reading of rural development challenges. To mention just a few such critiques: de Haan asks whether the livelihood approach should

be considered as a neo-liberal project … [and states that] it tended to focus much more on opportunities than on constraints, more on actor’s agency than on structure, more on neutral strategies than on failed access due to conflicts and inequalities in power.

de haan 2012: 348
Ramakumar (2008: 5) argues that ‘livelihood studies lost their politics’. With reference to key authors that are often mentioned in the genealogy of the Sustainable Livelihoods Framework, such as Freeman or Ellis (Freeman, Ellis, and Allison 2004; Ellis 1999), he shows how a more narrow analytical focus on poor people’s agency and the perceived need to diversify their livelihood options led to a normative attention to strengthening their agency. This strengthening became a key mandate for states, ngos and donors. However, as is the case in the Sindh province, the powerful role of these actors in actually creating livelihood crises escaped attention. As a consequence, livelihoods research became ‘a handmaiden of neo-liberal policies across the developing world’ (Ramakumar 2008: 4). Ramakumar therefore demands that ‘any meaningful livelihoods framework has to necessarily question the policies that cause crisis in the first place’ (Ramakumar 2008: 5).
In his search for theoretical entry points to understand everyday people’s agency and ‘fields of practice’ in contexts of violent conflict, Raeymaekers (2014) does deliberately not select a livelihoods approach, arguing that

instead of bridging the gap between micro- and wider conflict economies, this ‘sustainable livelihoods approach’ forestalled a more diversified understanding of the mediation between individual coping mechanisms and overarching mediating structures, not because these links are impossible to verify … but because of the dominant requirement of the international aid structure to reduce political problems to negative market externalities.

raeymaekers 2014: 33
Finally, Green and Hulme’s (2005) critique hints at the same disconnect:

The widely used livelihoods approach … as a ‘toolkit’ for understanding how households, assumed to be units of production and consumption, come to have income deficits is useful for helping outsiders grasp the different components of income that rural people have to pull together in order to make a living of sorts. It is less able to grasp the external influences on these disparate components, the extent to which rural dwellers are embedded in regional and transnational economies and the dynamism and often fragmentary nature of ‘households’ as loci of accommodation, consumption and cooperation.

green and hulme 2005: 868
Indeed, these and other critiques touch upon important limitations of how the livelihoods approach is being used by some researchers interested in the fate of marginalised people. In summary, most critics disqualify the livelihoods approach due to, first, its analytical insensitiveness vis-à-vis issues of power relations; second, its normative embeddedness in the zeitgeist (de Haan 2012: 346) of neo-liberalism; and third, its close and often uncritical affiliation to, and collaboration with, the development industry or ‘international aid structure’.

But I venture to argue that these critics, though highlighting important concerns, risk throwing out the baby with the bath water. They gauge their reading of the livelihoods approach vis-à-vis one specific understanding of what political – and for that matter apolitical – means, of how power and politics is to be understood and how not. Again, facets of this critique are crucial. But it also risks reducing the livelihoods approach to a rather narrow interpretation. Thus, it falls short of its potential as a critical instrument for analysing social relations and power (as indeed highlighted by de Haan, 2012: 345). As a matter of fact, some of the early applications of the livelihoods approach were already rather sensitive to broader questions of power relations and political economy. Ashley et al. (2003) for example explicitly addressed issues of social exclusion in their operationalisation of the livelihoods approach for rural studies in India; and Start and Johnson (2004) – also using insights from India – discussed the relations between livelihoods and broader questions of the social relations of production and exchange. But it is important to take the critique serious. This requires renewed attention to underlying assumptions, and thus ontology. After all, the study on Sindh also introduced class, but missed some crucially important facets of class relations.

Disentangling Underlying Ontologies

My core argument is that the claim for the livelihoods approach’s blindness to power relations is important, but misses the point; what matters is how these power relations are understood. Scoones has already noted:

One of the recurrent criticisms of livelihood approaches is that they ignore politics and power. But this is not strictly true. Livelihoods approaches encompass a broad church, and there has been some important work that has elaborated what is meant, in different variants of different frameworks, by ‘transforming structures and process’, ‘policies, institutions and processes’, ‘mediating institutions and organisations’, ‘sustainable livelihoods governance’ or ‘drivers of change’ … These reflections have addressed the social and political structures and processes that influence livelihood choices. Power, politics and social difference – and the governance implications of these – have been central to these concerns … Unfortunately, though, such debates remained at the margins.

scoones 2009: 11
In addition, I suggest that it is the respective conceptualisations of power that need far more critical scrutiny. This would then allow for a more differentiated interpretation of the livelihoods approach’s intersection with issues of power relations, the zeitgeist of neo-liberalism and its bond with development practice.

The name of Henry Bernstein appears rarely in the genealogies of the Sustainable Livelihood Framework. Its main parentage is usually seen as being ideologically or practically linked, for example, through research contracts, to the uk’s Department for International Development (dfid) at the time of Tony Blair’s ‘Third Way’ government (de Haan 2012: 346; Geiser et al. 2011: 258; Scoones 2015: 35f). Above, I have already referred to Freeman et al. (2004) and Ellis (1999), and its core founders of course include Chambers and Conway (Chambers 1983; Chambers and Conway 1992; for further details of these ancestors, see de Haan [2012] and Scoones [2009; 2015]).

But it is with Bernstein that I shall attempt to argue my point. In 1992, he and his co-authors edited a book entitled Rural Livelihoods: Crises and Responses (Bernstein, Crow, and Johnson 1992). One important intention of the book was to show that ‘[the] different ways in which land and labour are combined in farming is fundamental in understanding rural livelihoods, and whose livelihoods’ (Bernstein et al. 1992: 24; italics mine), indicating the authors’ background in critical agrarian studies. In his own chapter on ‘Poverty and the Poor’, Bernstein (1992) reflects on the basic positions from which researchers look at (rural) livelihoods. He uncovers dominant ontologies (though not using this notion) by pointing at the taken-for-granted views on reality, the unquestioned assumptions on how the world is, the (filtering) spectacles through which things out there are seen. Bernstein distinguishes, in an ideal-type manner, between a residual and a relational perspective:

The residual approach views poverty as a consequence of being ‘left out’ of processes of development, on the assumption that development brings economic growth which, sooner or later, raises everybody’s income. This is termed the ‘trickle down’ effect: that the benefits of growth trickle down even to the poorest groups in society in the form of increased opportunities to earn (more) income. The implication for development policy is to target the rural poor in order to integrate them into processes of development they have been excluded from. In practice, this typically means integrating them more deeply into markets and devoting more of their resources and energies to producing goods for sale.

bernstein 1992: 24
That, in a nutshell, is how the world out there is seen by a section of researchers and practitioners involved in rural development. Today, this section encompasses specifically the dominant neo-liberal school that mainstreamed most of the contemporary development debate and practice. Thus, I label this position in the following as ‘residual-mainstream’.
In contrast, Bernstein (1992) characterises a relational take on the world in these words:

Relational approaches investigate the causes of rural poverty in terms of social relations of production and reproduction, of property and power, that characterize certain kinds of development … A relational approach thus asks rather different questions: are some poor because others are rich (and vice versa)? What are the mechanisms that generate both wealth and poverty as two sides of the same coin of (capitalist) development?

bernstein 1992: 24
In a relational approach, poverty results not from the persistence of subsistence production and lack of market integration (as in the residual approach), but from the persistence of exploitative social relations. To quote Bebbington (2007: 793): ‘Poverty endures because of the social relationships and structures within which particular social groups are embedded … Chronic poverty is a socio-political relationship rather than a condition of assetlessness’.

So far Bernstein’s typology of the residual and the relational (see also Geiser 2014) which, though schematic, helps and invites us to ask for the underlying assumptions within the livelihoods approach. At the same time, though, it invites us to reflect on the respective assumptions in the critiques of the livelihoods approach too. I shall therefore now discuss the links between Bernstein’s notion of residual/relational and the debates around the livelihoods approach. I distinguish in the following between those researchers/practitioners that stick to the assumptions of the mainstreamed residual perspective in the next section and those that stick to a more orthodox reading of the relational one in the subsequent section. Thereafter, I shall then venture to review livelihoods-oriented studies that also emphasise the importance of the relational, but complicate the dichotomy of residual and relational. As I will show, they give refined attention to agency (the focus of the residual-mainstream) and structure (the focus of the orthodox relational) – which then, in turn, complicates the thinking about development interventions. But more on this later.

The Livelihoods Approach of the ‘residual mainstream’

I propose that most critics of the livelihoods approach address its use by what can be called, following Bernstein, the residual approach. Researchers wearing these ontological glasses are indeed puzzled by the inability of poor people or households to make better use of the opportunities offered, in principle, by expanding markets. They are impressed by the poor’s skills in coping and surviving, even in difficult situations, and highlight their coping strategies around income diversification, in cash or in kind. The famous livelihood ‘asset pentagon’ (dfid 2000) helps them to better grasp these skills and identify avenues for supportive interventions. Supporting the poor’s skills in coping and diversifying people’s incomes is seen as development agents’ crucial task. In line with contemporary discourses, these agents are not only the states and their development agencies, but also ngos (as is the core idea of the report on Sindh under examination above as well), and increasingly the private sector. Thus, we find – to use the development industry’s language – livelihood-oriented public–private partnerships (ppps) around value chains, for example. After all,

the assumption is that the end point, with agriculture as a business, driven by entrepreneurship and vibrant markets, linked to a burgeoning urban economy, is the ideal to strive for.

scoones 2009: 15
The objective of development is clear in the mainstreamed residual approach, and this in turn informs what livelihoods research needs to uncover.
There are such examples from Pakistan too. The World Bank observes a lack of growth, which stems from ‘many of the underlying challenges that have held Pakistan back in making progress … corruption, violence, vested interests and bureaucratic inertia (World Bank 2014: 1), and from the poor’s lack of assets (skills, credit, etc.) that would enable them to come out of poverty. Thus, the Bank advocates greater growth while at the same time taking care of the poor. More concretely in the case of Pakistan, this translates into a need for

strengthening the business environment … to improve competitiveness and expand investment, improve productivity of farms and businesses, and make cities more growth friendly to create productive and better jobs (especially for youth and women) … [supplemented with the strengthening of] safety nets.

World Bank 2014: iii
My argument is that this mainstream-residual perspective on livelihoods is not per se power-blind, but has a specific understanding of power. To illustrate my point it is interesting to interlink with de Haan’s (2012: 350) recollection of older discussions on power that proposed to differentiate, among others, between ‘power over’ (i.e. to control something), ‘power with’ (‘a collective power based on mutual support’), and ‘power to’ (a creative power that ‘allows actors to exercise their agency’). One could argue that it is ‘power to’ that is central in the residual’s take on livelihoods. Following its neo-liberal ideology – or ontology for that matter – the strengthening of individual and household capacities to interlink with wider market forces is central and thus rural producers are to be ‘empowered’ to become innovative entrepreneurs. ‘Power with’ has become important as well, as the intensive engagement of this particular livelihoods approach with issues of participation, community-based organisations (cbos) or state decentralisation indicate (Geiser and Rist 2009). However, ‘power over’ is not on the radar, and thus critical relational issues, including the role of the state or ngos, are not part of the analysis. Therefore, for this group of livelihoods researchers, the approach is not apolitical at all, but indeed politicised. Even the World Bank realises that the poor

face specific governance problems in rural areas, such as deeply entrenched political and social structures, that are often linked to unequal access to land, which perpetuates severe inequalities and can lead to violent local conflicts.

World Bank 2007: 245
Others, though, do not agree, as they have a very different take on power and politics. This is shown in the next section.

The Livelihoods Approach of the (Orthodox) Relational Perspective

The ontological filter of what I call the orthodox relational perspective is more sensitive for issues of structures. The sufferings of the ‘working class’ are the result of the structure of production, which in turn is formed by those in control of the means of production. Linking up with the classification of power recalled by de Haan (2012: 350) above, ‘power to’ is conceptualised in aggregate terms. It is ‘the party’, the ‘progressive forces’, or in cases where ‘the party’ is in control, ‘the state’ (in singular), that has the mandate and the power to mobilise the rural masses. After all, and as in the case of the residual mainstream, the thelos of development (read: progress) is clear. And, by default, this reading is per se critical of the neo-liberal zeitgeist and its close relation to the development industry. No wonder ngos are seen as agents of capitalism. The poor’s own ‘power to’ is, however, perceived as extremely limited, and thus, to change their fate, the relationship between classes needs to be fundamentally changed. Therefore, ‘power over’ is central. It is through the mobilisation of ‘the peasants’ that exploitative social relations are to be overthrown, in view of creating a new society that ensures non-exploitative social relations. Feudalism is often an issue mentioned in the context of Sindh, for example, ‘a large segment of population remains poor, downtrodden and marginalised … due to large landholdings by the aristocracy-cum-political class’ (News on Sunday, 27 October 2013). These power relations get manifested especially in the unequal distribution of land, in exploitative land tenure arrangements such as sharecropping, and the exploitation of agricultural labour. The National Peasant Coalition of Pakistan (npcp) argues that the

country’s political and administrative system is dominated by the powerful feudal lords … Sindh is a province characterized with high land concentration and dominance of feudal lords. Hundreds of thousands of peasants work on lands belonging to these feudal [lords].

npcp 2014
This structuralist relational analysis of root causes informs how the problem might be overcome. At the overall level, people need to take control of the state: ‘The new politics is not an “end of the state” but the affirmation of the state as an instrument of people’s power, people’s democracy and people’s empowerment’ (Tariq 2010: 4). In Sindh, there are general demands for the abolition of feudalism and calls for land reform. One major practice is to organise rallies and demonstrations. Several organisations in Sindh, ranging from political parties to concerned local groups, regularly mobilise their followers for such events (for reports on recent events of this kind, see Pakistan Observer, 6 March 2016; Dawn, 17 February 2016). During these events, speakers denounce feudalism and urgently demand that feudalism be abolished and land reforms implemented. These are indeed very important events, and they highlight the widespread level of political awareness among Pakistan’s critical civil society.

In actual practice, and beyond these demands for structural change, however, there are few details given of exactly how feudalism is to be abolished and land reform implemented, or who ‘the peasants’ are. At times, there is mention of the plot size that might be required for peasants to operate a viable entity (e.g. Ercelan et al. 2013: 3). Thus far I have come across no study that attempts to draw differentiated lessons from Pakistan’s earlier attempts at land reform in the 1950s and 1970s. Therefore, other concerned activists and organisations are rather critical about this kind of demand and action. Although they too recognised that the underlying problem is feudalism, they argue that events such as demonstrations remain ‘sloganeering’ (see also Akhtar 2012: 28) and that rallies hint at problems but seldom bring improvements to those they speak for.

I will come back to these criticisms in the next section. But first, a brief positioning of the critiques of the livelihoods approach and their blame of it being apolitical is due. I propose that most accusations of the apolitical emerge from a more orthodox relational perspective characterised by a rather static differentiation of the social universe. This is not only found in academic research (see below), but at the level of activism in Sindh as well. Many leftists engaged in rural issues in Pakistan, and in Sindh in particular, are deeply critical of ngos, which are regarded as representing the neo-liberal (or residual mainstream) approach to rural issues. The following quote is just one such statement among many:

[Civil society] has become ‘projectised’ by external donor funding and is prey to whims and fads of development aid … ngos are busy in welfare or service provision, and have yet to perform a transformatory role for society. They are also bound by donor agendas, lack of funds and cannot make an impact.

sattar 2012: unpaginated
Indeed, ngos are often seen as being a proxy for critically engaging with the livelihoods approach. This practice of accusing ngos of being apolitical has also become a field of intensive debate in itself within academia. I just hint at the notions of ‘participation as a new tyranny’ (Cooke and Kothari 2001), and the more recent critique of participatory endeavours as being ‘post-political’ (Wilson and Swyngedouw 2014). Of interest for my argument here is that most of these studies are highly insightful and informative regarding the critical analysis of ngos, their participatory ventures or their engagements with the neo-liberal zeitgeist. However, they often remain rather vague and speculative – even simplistic – when it comes to discussions of how to go beyond critique. Mohan (2001: 163–164) for example demands a ‘moving beyond [through] transformative approaches’, for example through ‘radicalising hybridity’ – referring to the need to bridge the difference between local people and experts within a broader understanding of power relations. After the deconstruction of participatory approaches as apolitical, Swyngedow and Wilson (2014) advocate more

equality, solidarity and democracy, held together by the fidelity to the conviction that these can be realised geographically through continuous and sustained political struggle. The realisation of the practices of these principles involves the self-organisation and self-management of the people, and will eliminate the state as the principle organiser of political life.

swyngedouw and wilson 2014: 304

Critical Relational Debates on Livelihoods

A brief summary of points discussed so far: to underpin my argument that the how of linking analysis with action in the livelihoods approach deserves more attention than just blaming it as apolitical, I introduced in the previous sections the residual as well as the relational take on rural issues, and I illustrated this around evidence from Pakistan. I argued that both these ontological positions in their mainstreamed or orthodox gaze address issues of power and politics. As a consequence of the residual ontology’s emphasis on ‘power to’, it stresses the agency of rural poor, and their lack of assets. As the Sindh report under examination writes, it is crucial ‘to support the ingenuity and resourcefulness, of both men and women to negotiate the challenges they are facing and will face’ (spdc 2015: 94). The relational perspective emphasises ‘power over’, and thus focuses its analytical take on structural dimensions. The definition of distinct, homogenous classes comes centre stage. Thus, the abolition of feudalism is, for example, an objective at intervention level. One emphasises ‘power to’, the other ‘power over’. Therefore, both positions highlight power-related aspects of why people stay poor.

However – and this leads me on to the next point in my argument – both neglect to address, in a differentiated manner, the relation between ‘power to’ (or agency for that matter) and ‘power over’. Thus, renewed attention to agency and structure is needed to overcome this shortcoming. Indeed, in recent decades, many researchers and development practitioners have – either explicitly or implicitly – engaged with a more critical and differentiated analysis of the relationship between diverse social groups in specific contexts, their agency and how their agency is influenced by wider political-administrative, economic and socio-cultural structures. They also analyse how these hindering or enabling structures come about, and are produced and shaped in turn by specific actors.

In doing this, they link up – again, implicitly or explicitly – with a range of debates within social sciences that inspired the sphere of development studies. A few hints: of early importance was the actor-oriented perspective by Long and Long (1992). Long and Long’s work in turn was inspired by Giddens’s conceptualisation of agency-structure through his structuration theory (Giddens 1984). Somehow, though, this particular theory has lost its lustre, and Scoones, for one, calls it ‘cumbersome’ (Scoones 2009: 17) without, however, explaining why. Other critical social science debates have unfolded around different takes on agency and structure, the most prominent of which are attempts to engage with Bourdieu’s concepts, for example, social capital. Thieme and Siegmann (2010) challenge the residual mainstream’s hype with social capital. They show that women’s livelihoods can suffer, and their vulnerability increase, from being incorporated in male-dominated social networks. Others build and expand on the notion of ‘institutions’ to grasp the relational. Steimann (2011) uses, among other things, the notion of legal pluralism to analyse how agro-pastoralists are affected by and cope with the post-socialist transformation in Kyrgyzstan. In their reviews of the livelihoods approach, both de Haan (2012: 349–352) and Scoones (2015: 46–60) add other references to these approaches. They also refer to studies emerging from more differentiated agrarian studies as well as to political ecology and its sub-schools. Indeed, there is an inspiring, lively, at times confusing but always engaging array of research taking place on how to analyse livelihoods. All these studies pay due attention to the centrality and complexity of how social relations work.

In the remainder of this section, I shall illustrate the implications such differentiated takes can have on our understanding of rural challenges. I specifically address this around the issues that I criticised in the study report on Sindh; that is, first, the ways in which the involved actors are categorised; second, how the social relations between these actors are addressed; and third, how the role of the state and ngos in poverty alleviation are discussed.

Categorising Rural People/Actors

While examining the Sindh report in the third section, I highlighted the challenge of stratifying the rural social universe. The report identified small farmers, communities, the state and ngos as actors. This stratification guided the understanding of the problems and potential of each group. Based on my own research, I then showed that the category ‘small farmer’ conceals and makes invisible crucial differences between farmers with little land and those who have a lot of land, and that it completely misses the haris or sharecroppers, who are often bonded labourers. This critique of the ‘small farmer’ category is received wisdom, but it needs reiterating again and again.

I also criticised the understanding of ‘the state’ as a homogeneous category and as external to small farmers’ problems. I showed that, on the contrary, unequal access to land and water is substantially caused by the de facto actions of proponents of the state. Another example is the ineffectiveness of the state’s Tenancy Act, which helps to reproduce the fate of the haris.

But the same challenge holds for classifications used in the more orthodox relational perspective as well. The stereotyped categorisation of classes as ‘peasants’, ‘workers’ or ‘landlords’ completely misses the enormous heterogeneity of social entities that form the social universe of Sindh, a heterogeneity that complicates social relations as well. It specifically misses the diversity of interests that might exist within and between each of these social categories. As Herring and Agarwala (2006) put it more abstractly:

Not all who are objectively members of a class may recognize that position; not all who recognize their location in a class structure will find that particular dimension of inequality most salient, or most amenable to change; not all who seek to alter the terms of their class position will find sufficient colleagues to make collective action feasible; and not all class-based collective action will be effective: much will be suppressed, bought off, tactically flawed, or ignored by political actors with alternative support bases.

herring and agarwala 2006: 333

Conceptualising and Addressing Social Relations

O’Laughlin (2004) argued that

class, not as an institutional context variable, but as a relational concept, is absent from the discourse of livelihoods. Accordingly, political space is very limited – focusing mainly on ‘empowering’ the poor, without being clear about how this process takes place or who might be ‘disempowered’ for it to occur. A more explicit theorisation of politics, power and social difference is thus required.

o’laughlin 2004 quoted in scoones 2009: 186
This critique speaks to the residual reading of rural livelihoods discussed above, but I suggest that it also holds for the (more structuralist and orthodox) relational perspective. As I illustrated above using examples from Pakistan, both tend to engage with rather static social categories (the ‘small farmers’, the ‘peasants’). They also base their take on the relations between these categories on what the respective ontologies prescribe (either functionalist relations between producers, the market and the state, or antagonistic relations between labour, capital and the state). What is missing from both perspectives is attention to the nitty-gritty of the relational (between a multitude of social groups) per se. This is a sphere that needs detailed empirical investigation. Only then can a more dynamic, process-oriented, constructivist and contingent understanding of social groups and their interactions emerge. I will illustrate this below around how the role of the state and of ngos vis-à-vis the ‘poor’ is discussed in Sindh.

The Role of the State and ngos

Many social activists and their organisations in Sindh share the relational mainstream’s framing of core challenges – feudalism, the plight of the haris, the need for land reform, and a critique of the state. But they want to go beyond rallies and sloganeering. In doing so, they venture to share some thoughts with the residual mainstream: peasants with a little land can produce little and, in addition, many lack adequate inputs (water, proper seed, implements, levelled fields, credit) to use their land productively. Some activists argue that these challenges would remain even after radical land reform.

Interestingly, several interviewees mentioned that some sections of ‘the state’ do provide laws and special schemes that do, in principle, address issues of poverty and inequality. However, many of these are either not implemented or are difficult for the marginalised to access. These activists then argue that what is ‘the need of the hour’ in this specific political-economic context, is not to just blame the state as such, but to demand the implementation of such laws and schemes from the state agencies involved. They also argue that people have the right and are entitled to access such state provisions. They thus engage in a critical interaction with these state institutions, based on a differentiated and in-depth analysis of their mandates as well as their actual performance.

The following is an example that emerged during my own research in Sindh. In 2008, the government of Sindh launched a project to distribute, through its Board of Revenue, ‘212,864 acres … land among landless haris, preferably amongst the women workers in almost all the districts of Sindh’ (Dawn, 6 October 2008). In all, 13,300 families were to benefit. The objective of the scheme was to ‘reduce poverty, access of landless to the basic livelihood sources and ultimately to ensure the food security’ (Legal Rights Forum [lrf] 2013: 5). Micro credit, skill training, and seed were to be provided in addition to land. Programme implementation was to be done through state-affiliated rural support programmes. The revenue officials then declared 85,199 acres as available for distribution, much less than initially foreseen. The project itself was implemented in two phases up to 2010.

A subsequent review of the scheme by the Legal Rights Forum (2013) lauds the government’s efforts, but finds, based on a very differentiated analysis of the scheme’s everyday nitty-gritties and with a clear focus on issues of social relations, severe problems in its implementation. Similarly, taking stock of the programme, Participatory Development Initiative (pdi) and Actionaid (2013: 6) report that 24% of the land grantees ‘are yet deprived of the access to the land granted to them’. In order to understand this anomaly, pdi and Actionaid carefully followed the scheme, and disentangled its various procedural steps. They found that in a first step, the available land was to be announced by the Board of Revenue of the Sindh province through the print media in order to seek applications. This first step thus already excluded many marginalised from even knowing about the scheme, as haris in particular have a low literacy rate.

Out of the applicants, ‘genuine landless hares’ were then to be identified by the authorities and given land. The review found that between 2008 and 2010 in total, 55,439 acres were distributed to 4867 women and 1266 men (pdi and Actionaid 2013: 6). One source argues that initially the size of allotted land was to depend on land quality (16 acres in irrigated areas, 25 in non-irrigated areas; Dawn 13 January 2014), but that smaller plots were actually distributed. Much of the distributed land was uncultivable, for example, due to salinity, being unlevelled, waterlogged, a graveyard or disputed. The initial identification of allottees was to be done through transparent and open Kacheries (i.e. open sessions or meetings where applications for land are scrutinised by state officials and the concerned public), but were often done at the ‘residence of the political influential’. Many potential beneficiaries were even not aware of these Kacheries. The reviews also claim a lack of ‘participation of civil society and local representation in identification and distribution of land’ (lrf 2013: 7).

Once haris were allotted land, they had to visit the Board of Revenue’s office at the district headquarters to request a document called an ‘allotment order’; again, many found these procedures difficult. Also, those entitled to land (the allottees) had to pay an initial ‘deposit’ of pkr100 and then an annual ‘documentation fee’ of pkr100 for the coming 15 years. The report claims that many were not aware of these conditions and thus lost their entitlements.

Next, allottees were to receive ‘support economic packages’ to enable cultivation, mainly through rural support programmes. Many of these packages came late because of conflicts around the criteria of eligibility. The rural support programmes followed the ‘Poverty Score Card System’ designed by the World Bank (Participatory Development Initiatives [pdi] and Actionaid 2013: 10), whilst the Sindh government used another system. Finally, allottees were meant to receive proper ownership documents, notably the crucial Form No. vii (lrf 2013: 11). For this, however, they required the ‘allotment order’ mentioned above as well as other documents, again opening a vast field of conflict, (mis)information and contestation.

So the question now is how one should intervene in this complex situation after such an analysis based on a differentiated understanding of both the ‘small farmers/peasants’, and ‘the state’. I generally argue that differentiated analyses of rural development challenges complicate the definition of development-oriented interventions. The above description suggests that the original land allocation scheme in Sindh was inspired by a residual ontology – in a nutshell, the poor were to be given assets by the state and related ngos. The differentiated reviews above, however, highlighted that this mainstreamed approach could founder on the iceberg of social relations.

It is interesting to analyse how activists in Sindh cope with this challenge when faced with the question of how to translate such critical insights into concrete interventions. In the case of the scheme under discussion, some smaller local organisations supported the allottees by helping them through the complex administrative procedures and by translating official government forms into local languages. They also supported them directly through ‘better support packages’. But they also agitate by means of rallies through a local network. For this, they develop close links to the media. They even support allottees in filing suits at the courts in charge. Of special importance is their support to allottees in obtaining the aforementioned Form No. vii, as well as proper receipts for the pkr100 they pay annually. They also support court cases around land that was allocated but not suitable for cultivation.

Indeed, these are just brief references to several very active local organisations in Sindh that engage specifically with the plight of haris and very small farmers-cum-labourers. These organisations go beyond mainstreamed analyses (whether residual or relational), and thus have to innovate, and engage in, rather unconventional practices – supporting court cases; critically challenging the Board of Revenue; mobilising the police to raid landlords that imprisoned hari families. Some organisations also try hard to mobilise haris in interest groups. In order to avoid the residual’s (to stay with this terminology) trap of ‘community-based organisations’ (cbos) and their informality, the National Trade Union Federation (ntuf) is endeavouring to register hari groups as formal trade unions. This in turn requires complex struggles to amend the legal frameworks, as the respective legislation of the Sindh government – the Sindh Industrial Relations Act – foresees the modality of trade unions for industrial labour only. ntuf recently succeeded in this, opening a potentially new resource for the struggle of haris (ntuf 2014; Daily Times, 31 January 2016).

Again, these are just examples of activities that go beyond ‘sloganeering’ or the distribution of assets. On the side of the analysis, they are based on in-depth understandings of the actual and everyday working of social relations. On the side of the definition of interventions, they are innovatively addressing social relations. But this is not simple, to say the least, and requires long-term involvement and dedication, and at least some funds. Such funds are, of course, not coming from the state.

This brings me, briefly, to the social relations in which these local-level activists and their non-governmental organisations are enmeshed themselves – social relations that, by the end of the day, can impinge on the faith of the poor as well. I learned through my interviews that most international donors will not provide funding for the above kinds of rather political activities, as such actions are considered too sensitive. Some larger Pakistani non-governmental organisations (more precisely, some concerned staff within these organisations) try at least to squeeze some support for haris into their more mainstreamed projects that do receive international funding. Such mainstreamed and ‘politically correct’ projects include food security, climate change adaptation, or gender mainstreaming. But I found that many local activist organisations are reluctant to enter into these arrangements, mainly because they are usually limited/projectised to a few years only, and feared distracting attention from core relational issues. These activists insist that issues of social relations are central, complex and in need of explicit and dedicated long-term commitment. To recall – the study summarised in the second section, dealing with ‘climate change’, did not consider the fate of haris as well.

For a Critical Relational Livelihoods Perspective: Conclusion

The above examples illustrate that how power and politics are conceptualised in a livelihoods approach, and how power and politics are then addressed in the sphere of interventions is crucial. They also illustrate that more differentiated understandings, especially regarding the stratification of the rural populace, the social relations between them, and the role of the state and different types of ngos, complicate clear-cut answers on how to intervene for a more ‘constructive engagement’ (Raeymaekers 2014: 41). It is not that easy to identify ways out of power imbalances and their consequences for the everyday lives of millions of people without falling back on the clear-cut templates of either the residual mainstream or the relational orthodoxies (see e.g. the quote from Swyngedow and Wilson [2014] further above), and accusing each other of being apolitical or sloganeers.

It is through the engagement with more contemporary and critical social science theories beyond individualistic or structuralist positions that a much more dynamic understanding of ‘[access] to livelihood opportunities [which] is governed by social relations, institutions and organizations, and [which] includes power as an important explanatory variable’ (de Haan and Zoomers 2005: 44) becomes possible. These approaches indeed emphasise power, but understand the emergence and everyday working of power in much more relational and dynamic ways (see especially Harriss 2007; Mosse 2010). This necessitates nuanced and contextualised analysis, qualitative for that matter. Questionnaires can show class as a category of people, but not as a relational concept. Interestingly, a lot of such nuanced analyses emerge from outside academia, out of the experiences of social activists like the ones I had the privilege to meet in Sindh.

Differentiation, though, implies complication too. Power is understood in much more nuanced ways; the zeitgeist is differentiated as well; and the issue of ‘so what’ after the critical-radical-differentiated analyses becomes a field of contestation of its own. Simply blaming the state or the development industry does not help any longer. In conclusion, I fully agree with Scoones (2009: 14) who states that ‘what is needed is a re-energising of the livelihoods perspectives’, but this can complicate the livelihoods of researchers and development practitioners themselves. After all, which of the big players in the development industry is ready to finance a critical analysis of the fate of haris, or court cases by haris against landlords and the state?

Acknowledgments

I am grateful to Nicolas Martin, Hussain Bux Mallah and Muhammad Ismail Kumbhar for detailed readings of the paper and for many constructive comments. Responsibility with remaining shortcomings, of course, rests entirely with me.

References

  • ACCESS (2015), Livelihood Asia Summit 2015. http://www.accessdev.org/lh-livelihoods-asia.php[Accessed February 2016].

  • Akhtar A.S. (2012), 21st Century Socialism in Pakistan? Economic & Political Weekly (EPW), November 10: 2729.

  • Ashley C., D. Start, R. Slater, and P. Deshingkar (2003), Understanding Livelihoods in Rural India: Diversity, Change and Exclusion.London: Overseas Development Institute.

    • Search Google Scholar
    • Export Citation
  • Bebbington A. (2007), Social Movements and the Politicization of Chronic Poverty. Development and Change, 38 (5): 793818.

  • Bernstein H. (1992), Poverty and the Poor. In: H. Bernstein, B. Crow and H. Johnson (Eds.), Rural Livelihoods: Crises and Responses. Oxford: Oxford University Press, 1326.

    • Search Google Scholar
    • Export Citation
  • Bernstein H., B. Crow, and H. Johnson (Eds.) (1992), Rural Livelihoods: Crises and Responses. Oxford: Oxford University Press.

  • Budhani A., and H. Gazdar (2011), Land Rights and the Indus Flood 2010–2011: Rapid Assessment and Policy Review. Oxfam Research Report. Islamabad: Oxfam.

    • Search Google Scholar
    • Export Citation
  • Bureau of Statistics (2011), Development Statistics of Sindh 2011. Planning and Development Department, Bureau of Statistics. Karachi: Government of Sindh.

    • Search Google Scholar
    • Export Citation
  • Chambers R. (1983), Rural Development: Putting the Last First. Harlow: Longman.

  • Chambers R., and G. Conway (1992), Sustainable Rural Livelihoods: Practical Concepts for the 21st Century. IDS Discussion Paper 296. Brighton: Institute for Development Studies.

    • Search Google Scholar
    • Export Citation
  • Cooke B., and U. Kothari (Eds.) (2001), Participation: The New Tyranny? London: Zed Books.

  • Daily Times (2016), Sindh Industrial Relations Act: Peasants Await Implementation. 31 January.

    • Export Citation
  • Dawn (2008), Criteria for Land Reforms in Sindh.6 October.

    • Export Citation
  • Dawn (2014), Land Distribution Among Peasants Suspended. 13 January.

    • Export Citation
  • Dawn (2016), Peasants Demand Land Reforms. 17 February.

    • Export Citation
  • de Haan L. (2012), The Livelihood Approach: A Critical Exploration. Erdkunde,66 (4): 345357.

  • de Haan L., and A. Zoomers (2005), Exploring the Frontier of Livelihoods Research. Development and Change, 36 (1): 2747.

  • Department for International Development (DFID) (2000), Sustainable Livelihoods Guidance Sheets. London: DFID.

  • Ellis F. (1999), Rural Livelihood Diversity in Developing Countries: Evidence and Policy Implications. Natural Resources Perspectives 40. London: Overseas Development Institute.

    • Search Google Scholar
    • Export Citation
  • Ercelan A., K. Ali, and M. Shah (2013), Land for Food Sovereignty in Pakistan. Discussion Note PILER. Karachi: Pakistan Institute of Labour Education and Research.

    • Search Google Scholar
    • Export Citation
  • Food and Agriculture Organisation (FAO) (2016), Increase the Resilience of Livelihoods to Threats and Crises. http://www.fao.org/about/what-we-do/so5/en/[Accessed February 2016].

    • Search Google Scholar
    • Export Citation
  • Freeman H., F. Ellis, and E. Allison (2004), Livelihoods and Rural Poverty Reduction in Kenya. Development Policy Review,22 (2): 147171.

    • Search Google Scholar
    • Export Citation
  • Geiser U. (2014), Conceptualising ‘Contested Development’ – from Grand Narratives to the Nitty-gritty of the Everyday. In: S.R. Sharma, B. Upreti, P. Manandhar and M. Sapkota (Eds.), Contested Development in Nepal: Experiences and Reflections. Kathmandu: School of Arts Kathmandu University and Nepal Centre for Contemporary Research, 125.

    • Search Google Scholar
    • Export Citation
  • Geiser U., and S. Rist (Eds.) (2009),Decentralisation Meets Local Complexity: Local Struggles, State Decentralisation and Access to Natural Resources in South Asia and Latin America. Perspectives of the Swiss National Centre of Competence in Research (NCCR) North–south, University of Bern vol. 4. Bern: Geographica Bernensia.

    • Search Google Scholar
    • Export Citation
  • Geiser U., U. Müller-Böker, B. Shabaz, B. Steimann, and S. Thieme (2011), Towards an Analytical Livelihoods Perspective in Critical Development Research. In: U. Wiesmann and H. Hurni (Eds.), Research for Sustainable Development: Foundations, Experiences, and Perspectives, vol. 6. Bern: Geographica Bernensia, 257271.

    • Search Google Scholar
    • Export Citation
  • Giddens A. (1984), The Constitution of Society: An Outline of the Theory of Structuration. Cambridge: Polity Press.

  • Government of Sindh (1950), The Sindh Tenancy Act 1950 (Sindh Act No. 20 of 1950). Karachi: Law Department Government of Sindh.

    • Export Citation
  • Green M., and D. Hulme (2005), From Correlates and Characteristics to Causes: Thinking About Poverty from a Chronic Poverty Perspective. World Development,33 (6): 867879.

    • Search Google Scholar
    • Export Citation
  • Green Rural Development Organisation GRDO (n.d.), Green Rural Development Organisation. http://grdo.org.pk/index.php/accomplishments/[Accessed March 2014].

    • Search Google Scholar
    • Export Citation
  • Harriss J. (2007), Bringing Politics Back Into Poverty Analysis: Why Understanding of Social Relations Matters More for Policy on Chronic Poverty than Measurement. Q-Squared Working Paper 34. Burnaby: Simon Fraser University.

    • Search Google Scholar
    • Export Citation
  • Hasnain T., and T. Marri (2015), Peasants’ Rights and Issues in Sindh: An Appraisal of Sindh Tenancy Act 1950. Karachi: National Peasant Coalition of Pakistan.

    • Search Google Scholar
    • Export Citation
  • Herring R.J., and R. Agarwala (2006), Introduction – Restoring Agency to Class: Puzzles from the Subcontinent. Critical Asian Studies, 38 (4): 323356.

    • Search Google Scholar
    • Export Citation
  • Human Rights Commission of Pakistan (HRCP) (2015), State of Human Rights in 2014. Lahore: Human Rights Commission of Pakistan.

  • Hussain G., and A. Mohyuddin (2014), Multiple Conventions of Unpaid Labour (begar) in Sindh Rice Belt: Dualistic Resistance and Cooperation in Sharecropping Arrangements. Science International,26 (1): 425431.

    • Search Google Scholar
    • Export Citation
  • Hussein M.H., A. Saleemi, S. Malik, and S. Hussain (2004), Bonded Labour in Agriculture: A Rapid Assessment in Sindh and Balochistan, Pakistan. ILO Working Paper 26. Geneva: International Labour Organisation.

    • Search Google Scholar
    • Export Citation
  • Indus Consortium (2015), Water Governance in Punjab and Sindh: A Research Study 2014–15.Islamabad: Indus Consortium.

  • Keely J., and I. Scoones (2000), Understanding Environmental Policy Processes: A Review.IDS Working Paper 89. Brighton: Institute of Development Studies.

    • Search Google Scholar
    • Export Citation
  • Leach M., G. Bloom, A. Ely, P. Nightingale, I. Scoones, E. Shah, and A. Smith (2007), Understanding Governance: Pathways to Sustainability. STEPS Working Paper 2. Brighton: STEPS Centre Institute of Development Studies.

    • Search Google Scholar
    • Export Citation
  • Legal Rights Forum (LRF) (2013), Women Public Land Grantees, Government of Sindh. Study Report. Karachi: Legal Rights Forum/Actionaid.

  • Lieten K., and J. Breman (2002), A Pro-Poor Development Project in Rural Pakistan: An Academic Analysis and a Non-Intervention.Journal of Agrarian Change, 2 (3): 331355.

    • Search Google Scholar
    • Export Citation
  • Long N., and A. Long, (Eds.) (1992), Battlefields of Knowledge: The Interlocking of Theory and Practice in Social Research and Development. London: Routledge.

    • Search Google Scholar
    • Export Citation
  • Mohan G. (2001), Beyond Participation: Strategies for Deeper Empowerment. In: B. Cooke and U. Kothari (Eds.), Participation: The New Tyranny? London: Zed Books, 153167.

    • Search Google Scholar
    • Export Citation
  • Mosse D. (2010), A Relational Approach to Durable Poverty, Inequality and Power. Journal of Development Studies,46 (7): 11561178.

  • National Peasant Coalition of Pakistan (NPCP) (2014), Pakistan’s First Peoples’ Tribunal on Land Rights in Sindh. Karachi: National Peasant Coalition of Pakistan (NPCP).

    • Search Google Scholar
    • Export Citation
  • National Trade Union Federation (NTUF) (2014), Registration of Hari Trade Unions Termed Historic Step, Thousands of Haris Participated. Socialist Pakistan News (Karachi), 18 March.

    • Search Google Scholar
    • Export Citation
  • News on Sunday (2013), Rights Activists Demand Land Reforms in Pakistan. 27 October 2013.

  • O’Laughlin B. (2004), Book Reviews. Development and Change, 35(2): 385403.

  • Pakistan Observer (2016), Sindh People’s Caravan Welcomed in Badin. 6 March.

    • Export Citation
  • Participatory Development Initiatives (PDI) and Actionaid (2013),Consultative Workshop on Land Distribution among Women in Sindh. Karachi: Participatory Development Initiatives.

    • Search Google Scholar
    • Export Citation
  • Raeymaekers T. (2014), The Social Geography of Violent Conflict. In: T. Bruck, P. Justino and P. Verwimp (Eds.), Micro-Level Perspective on the Dynamics of Conflict, Violence and Development. Oxford: Oxford University Press.

    • Search Google Scholar
    • Export Citation
  • Ramakumar R. (2008), Re-Politicising Livelihood Research. Keynote Lecture, International Conference on Research for Development (ICRD), University of Bern, Switzerland, 2–4 July.

    • Export Citation
  • Sattar N. (2012), A Divided Civil Society. Dawn, 6 May http://www.dawn.com/news/716195/[Accessed June 2015].

  • Scoones I. (2015), Sustainable Livelihoods and Rural Development. Winnipeg: Fernwood Publishing.

  • Scoones I. (2009), Livelihoods Perspectives and Rural Development. Journal of Peasant Studies, 36 (1): 171196.

  • Shahbaz B., U. Geiser, and A. Suleri (Eds.) (2013), Forests, Livelihoods and Power Relations in North-West Pakistan. Lahore: Saang-e-Meel Publishers.

    • Search Google Scholar
    • Export Citation
  • Shaikh M.A. (2001), Report on Agriculture in Sindh: Issues and Options. Karachi: Shaheed Zulfikar Ali Bhutto Institute of Science and Technology.

    • Search Google Scholar
    • Export Citation
  • Social Policy and Development Centre (SPDC) (2015), Gender and Social Vulnerability to Climate Change: A Study of Disaster Prone Areas in Sindh. Karachi: Social Policy and Development Centre.

    • Search Google Scholar
    • Export Citation
  • Start D., and C. Johnson (2004), Livelihood Options? The Political Economy of Access, Opportunity and Diversification. ODI Working Paper 233. London: Overseas Development Institute.

    • Search Google Scholar
    • Export Citation
  • Steimann B. (2011), Making a Living in Uncertainty: Agro-Pastoral Livelihoods and Institutional Transformations in Post-Socialist Rural Kyrgyzstan. Human Geography Series 26. Zurich: Zurich University.

    • Search Google Scholar
    • Export Citation
  • Swyngedouw E., and J. Wilson (2014), There is No Alternative. In: J. Wilson and E. Swyngedouw (Eds.), The Post-Political and Its Discontents. Spaces of Depoliticization, Spectres of Radical Politics. Edinburgh: Edinburgh University Press.

    • Search Google Scholar
    • Export Citation
  • Tariq F. (2010),South Asia: A Regional ‘New Politics’ Needed to Challenge Neoliberal Agenda. Paper presented at Conference on the Role of Market and State in Poverty Reduction and Improving Distributive Justice, 2–3 October. Colombo: Council of Social Democrats in Sri Lanka/Friedrich Ebert Stiftung.

    • Search Google Scholar
    • Export Citation
  • Thieme S., and K. Siegmann (2010), Coping on Women’s Back: Social Capital – Vulnerability Links Through a Gender Lens. Current Sociology, 58 (5): 715737.

    • Search Google Scholar
    • Export Citation
  • Wilson J., and E. Swyngedouw (Eds.) (2014), The Post-Political and Its Discontents. Spaces of Depoliticization, Spectres of Radical Politics. Edinburgh: Edinburgh University Press.

    • Search Google Scholar
    • Export Citation
  • World Bank (2007), World Development Report 2008: Agriculture for Development. Washington, DC: World Bank.

  • World Bank (2014), Country Partnership Strategy for the Islamic Republic of Pakistan for the Period FY 2105–19. Report No. 84645-PK. Pakistan Country Management Unit South Asia Region. Washington, DC: World Bank.

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.