This paper examines the implications of applying a sustainable livelihoods (SL) approach at the policy level. Policy is a broad concept that can be disaggregated into content, the process of policy formulation and policy implementation. Although generalization about all these three elements is, by its nature, a simplification, the very diversity of livelihoods makes it even more difficult to generalize about policy content.
There is a substantial body of work on the impact of policy on concepts related to SL, such as food security and poverty, which has a great deal of relevance to assessing policy impact and appropriate content. However, looking at policy through an SL lens requires greater emphasis on vulnerability and risk-proofing than is sometimes found.
In many countries, the policy process is essentially top down. Communities often have an extremely limited role to play, and even where acknowledgement is made of the importance of participatory processes, often these are used in an extractive rather than an empowering context. An SL-friendly policy process would indicate a much more active role for communities and civil-society organizations (CSOs). This could be in terms of involvement in formal planning processes, greater inclusion in the political process, particularly at a decentralized or local level, or through an increased role for civil-society organizations in lobbying. National NGOs and other CSOs will often need increased support, both technical and financial, to enable them to carry this out effectively.
For SL-friendly policy, the organizations involved, whether government, private sector, civil society or traditional authorities, must be competent, accountable, accessible and responsive to the situation of vulnerable groups.
Direct participation, lobbying and monitoring impact are all-important elements of an SL-friendly policy environment. However, the possibilities of popular participation in policy processes will vary according to the particular policy arena concerned. Lobbying and monitoring are likely to be the main communication links to government for macro policy, whereas there is more scope for participation and consultation in areas such as institutional reform and service provision.
Macroeconomic stability is an important element of economic sustainability, and planning tools such as public expenditure management have contributed in many countries to achieving improved economic management. These can and should be made supportive of an SL approach to policy.
The author feels that, although it is important to improve understanding about the linkages between policy and livelihoods in terms of content and impact, the greatest potential for achieving a sustainable improvement in livelihoods policy lies in an emphasis on increasing civil-society and stakeholder participation in the policy process. This has implications for the focus of external assistance, both technical and financial.
The purpose of this paper is to examine the implications of applying a sustainable livelihoods (SL) approach at the policy level, in particular to issues of poverty. It addresses three questions: (1) what characteristics of policy might be consistent with and supportive of the SL agenda; (2) what types of organization could play an active role in influencing policy; and (3) in particular how to address the challenge of trying to incorporate an SL approach into the policy process in order to develop effective channels of communication between poor households, especially in rural areas, and the central policy network.
The sustainable livelihoods approach was adopted by a number of agencies and organizations during the 1990s as an integrative framework for thinking about development issues, and in particular for addressing poverty.
The framework links the concepts of capability, equity and sustainability, each concept being seen as both good in itself and an end (Chambers & Conway 1992). The concepts are employed in both a social and an environmental context, thus, for example, sustainability is seen as encompassing such elements as the overexploitation of non-renewable resources and socio-economic resilience to external shocks.
The most generally quoted definition of livelihoods is that given by Chambers & Conway:
"a livelihood comprises the capabilities, assets (stores, resources, claims and access) and activities required for a means of living: a livelihood is sustainable which can cope with and recover from stress and shocks, maintain or enhance its capabilities and assets, and provide sustainable livelihood opportunities for the next generation: and which contributes net benefits to other livelihoods at the local and global levels and in the short and long term."
The ultimate objective of projects, programmes and policy is thus the promotion of sustainable livelihoods, particularly for the poorest and most vulnerable in society. The SL approach is people focused: the basic measure of success for projects and policies is the extent to which they enable individuals, households and communities to strengthen sustainable livelihoods for themselves.
It can be argued that the SL approach has evolved from thinking about poverty as a problem of lack of income, through to the basic needs approach, then to an emphasis on food security and vulnerability, and finally more recently to an approach to poverty programmes that focuses on the provision of health and education services by government. A sustainable livelihoods approach is likely to encompass elements of all these aspects, but focuses on capacities rather than needs, and on assets and strengths rather than weaknesses and constraints.
The precise frameworks and tools used by different agencies vary (see Carney et al. 1999). However, they all share the same basic concept of sustainable livelihoods, and use a framework that contains the following elements:
The different agencies have identified particular entry points for the SL approach that reflect both varying conceptual emphases and the agencies' own strengths and opportunities. Much of the practical use of the SL approach to date has been in terms of designing projects and programmes, a reflection of both the emphasis and the approach of the agencies that have adopted the approach (NGOs and bilateral and multilateral agencies) and the complexity of understanding the full implications of such a holistic approach for the policy process. It is this latter issue - the implications of the SL approach for the policy process that this paper seeks to address.
In the sense that policy can be seen as public-sector/government decision-making about both public resource allocation and the set of public institutions (rules of the game), this clearly has an impact on all the elements identified above. Perhaps the most immediate impact is on the stock of assets, or access to those assets, many of which are either public goods, such as infrastructure, or the outcome of the provision of public goods, such as education.
However, among the core principles of the SL approach is that it is people centred, responsive, participatory and dynamic. From this perspective the policy process is as important as its content, if not more so. The actual content of policy developed using the SL approach will tend to be specific to the individual situation. Certain areas might be expected to be given more prominence from an SL perspective, such as health, education and credit policy. However, it is in the way that policy evolves, the nature of the organizations involved in the policy process and the adaptability and responsiveness of policy to a changing environment, both physical and economic, in which the essence of the SL approach lies. Consequently this paper will concentrate as much on policy process as on content.
Given the limited experience of applying the SL approach at the policy level, the paper will also concentrate on identifying the characteristics of the policy process that are consistent with and supportive of the SL approach. Although overall the practical experience of using the SL approach at the policy level is limited, many of the tendencies and approaches to development that have been pursued over the past decade or so share similar concerns to the SL approach. These have resulted in changes in institutions and organizations that are supportive of and consistent with a SL approach. Where possible and relevant, examples of this are illustrated.2
The following issues or changes in paradigm have particular relevance to or significance for the SL approach:
The changing role of the State. There has been considerable debate over the impact of the changing role of the State on poverty and the access of the most vulnerable to services. In some countries the withdrawal of the State from provision of goods and services has restricted access of poorer households, particularly in more geographically remote areas, to reliable markets for both consumption goods and inputs for production. However, it also has to be acknowledged that in many countries, public-sector services benefited the rather better-off and those with good political connections, at the expense of the most vulnerable. The SL framework places more emphasis on supporting and enhancing capabilities, rather than simply meeting needs, as was the objective of old-style public-sector provision. Although these two approaches are not exclusive - and clearly food security, for example, is an important livelihood outcome - the redefined role of the State gives the potential for a more enabling environment for livelihood adaptation and sustainability.
Gender. This has been an important issue for the last decade. Aid organizations in particular have tried to ensure that their programmes be gender inclusive. An increasing number of countries have also introduced legislation to ensure the rights of women in inheritance, landholding, political structures, etc. From an SL perspective, improving women's access and participation is an important and indeed integral part of the process of achieving sustainable livelihoods. Because analysis is often carried out at the household level, it can be difficult to ensure gender sensitivity, but empowerment and participation of both genders is a primary objective and important element of the SL framework.
The sector approach. This is perhaps less obviously supportive of the SL approach than other approaches to development. It has been promoted as a response to the perceived shortcomings of project-led assistance. However, it has been criticized as ignoring the essential cross-sectoral nature of livelihoods and being inherently top down and non-participative. (Akroyd & Duncan 1998). In the past, addressing cross-sectoral problems has often proved to be organizationally intractable. The experience of food security policy and planning in many countries has been one of marginalization to the main sectoral concerns of the major line ministries. A more appropriate way forward may be to acknowledge the strengths of the sector approach, as a way of increasing the effectiveness of development aid and establishing realistic and efficient management and budgetary frameworks at the sectoral level, while exploring ways of making planning and policy processes more participatory and responsive. The recent Eritrean experience shows that participation can be integrated into an essentially sectoral process.
Good governance. This has increasingly become an important element in the aid relationship. The term good governance is used to mean "the mechanisms, processes and institutions through which citizens and groups articulate their interests, exercise their legal rights, meet their obligations and mediate their differences" (UNDP 1998a). As such, it is central to the objectives of ownership and empowerment that are critical to the SL framework. The rules that determine the conduct of public affairs also determine the access of individuals, households and communities and the set of incentives that determine livelihood strategies. Whereas sustainable livelihoods may not be totally determinate on the existence of good governance, where governance is predictable and robust, livelihoods become less vulnerable. The absence of good governance can exacerbate policy weaknesses. For example, in Nepal, a combination of traditional structures and authorities (including caste, bonded labour and complex and inequitable land tenure) and a public sector that is largely ineffective in assuring critical services have deprived both rural and urban populations of critical assets, and thus perpetuated extreme and widespread poverty.
Rights-based approaches. These have some overlap with good governance, though the emphasis here is on using the legal system rather than the political system to advance the position of disadvantaged groups in society. In practice, the establishment of rights within a legal framework is almost always the result of a political process. Many countries have signed declarations of rights within global conventions, but these rights have not subsequently been adopted by those countries' national legislative bodies. Even if adopted, the incorporation of rights into national legislation will have little impact unless a country has a progressive judiciary and a legal system that gives access to a broad spectrum of the population. Nonetheless, the establishment of appropriate rights at a national level can be an important tool in the pursuit of sustainable livelihoods.
Decentralization. This has been and still is being implemented in a number of African countries. The process of delegating implementation and, eventually, the budgeting and planning of line ministries to a regional and local level allows for much greater awareness of and responsiveness to local conditions and, in theory at least, accountability to local populations. From a policy perspective this should introduce a much greater two-way flow of information. As yet, experience is rather limited, but this trend should give rise to policy processes that are much more supportive of an SL approach.
The heading of this section can be interpreted at a number of different levels. The various interpretations of the SL approach itself can be seen as a spectrum with, at one end, a people-focused integrative way of addressing poverty issues, and at the other, a perception of a society as a dynamic and adaptive complex system that has to be analysed holistically (UNDP 1998b).
Policy itself can be analysed conceptually at a number of different levels. In its broadest sense, the term policy can be used to include projects, programmes, strategies, plans and their implementation and, in fact, every element of public or collective decision-making. Although it is a rather artificial simplification, policy can be divided into content and the process of policy formulation (in other words, the way in which that content is arrived at). The way in which policy is implemented can change the effective content of policy, either because policy interactions have not been fully understood or because the policy is subverted by those responsible for implementing it.
The nature of policy analysis, or advice on public decisions, is another important aspect of policy. For an external organization, such as a multilateral or bilateral agency, or an international NGO, the most important issue may be how adopting an SL approach affects the nature and context of the advice those agencies give their partners at the national or district level. Does it affect the nature of their interaction with other stakeholders?
Source: Adapted from Sutton 1999.
Figure 1 shows how different aspects of the policy process could be interconnected. In the figure, households have no direct impact on the policy process but are simply a source of information for policy analysts and bureaucrats/technicians. Policy affects them but they have little direct input in the process. Where government is democratically elected, or where policy-makers are in some ways dependent on households or communities, there may be linkages where those households or communities can have an impact on the policy process. This will be explored in more detail later.
Adopting an SL approach can change the content of the boxes in the figure. It may result in:
It may also change the type of information that is sought from the household or community level by policy analysts and bureaucrats/civil servants, both in terms of developing policy choices and in monitoring policy impact.
Finally, and perhaps most important, adopting an SL approach might lead to changes in the nature and direction of the linkages between the different agents in the policy process. Rather than being relatively passive providers of information, households would instead participate directly, or through civil-society organizations, in the determination of policy.
In other words, understanding the impact of policy on the capacity and opportunities to enhance and support sustainable livelihoods is an important element of analysing the SL framework. This can, in turn, lead to changes in policy content and implementation. However, the emphasis on participation and capacity-building that arises from an SL approach has implications for the policy formulation process, and this should, in turn, have implications for policy content.
Policy content is potentially a vast topic. The content of policy that is supportive of SL will be specific to given situations, countries and even districts within countries. Given the diversity of livelihoods within any one country, specific policies are likely to be more supportive of some livelihood strategies than of others. They may well have negative effects on some groups. Even where a major policy objective is the reduction of poverty and the improvement of livelihoods, there may be the need to make policy choices that are conflicting, favouring some livelihoods at the expense of others. This highlights the importance of the processes by which these decisions and choices are made, and the influence that different sections of the community can bring to bear on the policy process.
The role of multilateral and bilateral agencies and international NGOs in the policy process varies according to the level at which policy is formulated. In many cases, these organizations are stakeholders primarily in the sense that they provide funds and assistance to projects and programmes whose outcome is dependent on the policy environment.
Direct support to the process of policy formulation has tended to take one of a number of forms:
Adoption of the SL approach has implications for all these aspects, primarily in terms of the nature of the capacity-building supported, the specific areas where technical assistance and analysis are provided and the skills and orientation of the donor community and of the policy advisers and analysts engaged to assist national governments. The SL approach should provide a broader, more cross-sectoral perspective on policy issues. More fundamentally it should emphasize providing assistance to national governments in achieving a broader input into the policy process, increasing responsiveness and building the capacity of civil society to play an active role in policy determination.
It is easy to fall into the trap of considering policy as being formed only at the national level. In fact, policy is formed and implemented at a number of different levels, both nationally and internationally.
Internationally, organizations such as the World Trade Organization are important actors in determining conventions and institutions that affect the interaction of nations, through trade and through the rights and regulations they accept and incorporate in their national policies. These international-national linkages can be as important as micro-macro linkages in supporting and promoting sustainable livelihoods. International fora are usually more multisectoral in nature than national policy fora, and therefore, in one sense, they lend themselves to the incorporation of an SL agenda. However, the countries represented at these fora have internal diversity in interest. There is a need, in both North and South, for an improved analysis of the implications of an SL agenda. Civil-society organizations in the developed and developing world have an important role to play in lobbying at the national and international level to have SL concerns placed on the public agenda.
The policy focus is usually at the national level, where most decisions on legislation, regulation and resource allocation and public spending are made. This is gradually changing in many countries as greater levels of decentralization are being introduced. One of the most important challenges in incorporating an SL approach into the policy process is developing effective channels of communication between poor households, particularly in rural areas, and the central policy network. In part this may be a question of commitment, but even where commitment exists, the challenges for the policy-maker in responding to the complexity of livelihoods in a country is immense. This challenge may be somewhat smaller at the subnational and district level. The degree of homogeneity in livelihoods may be greater, and the geographical distance between policy-maker and communities may be reduced, but many of the key issues are the same.
Sectoral policy presents a different type of challenge to the SL analyst. The SL framework is, by its nature, intersectoral. Individuals and households adopt livelihood strategies from the whole range of possibilities open to them, and they do not restrict themselves to an individual sector. Yet sectoral policies and investment programmes focus on the activities of a number of restricted line ministries. This has often achieved improved sectoral management, but sometimes at the cost of a narrowly focused set of policy objectives. Where the policy context is predominately sectoral it can be difficult to achieve "joined-up" policy. In addition, line ministries often have much closer links with quite restrictively defined civil-society organizations (e.g. Ministries of Agriculture with farmers' unions). None of these facets of sectoral policy is inevitable, but it can mean that adopting an SL agenda, or policy objectives places additional demands on a limited technical capacity at the ministry level.
The rest of this paper explores in more detail the implications of the SL approach for development policy. It then examines what characteristics of the policy process might be consistent with and supportive of a sustainable livelihoods agenda. The types of organizations that could play an active role in influencing policy are discussed, along with the ways in which external agencies could build their capacity to undertake that role. Finally, the implications of the SL framework for public expenditure management are explored.
It could be argued that the approach taken in this paper is somewhat reductionist. Certainly it starts from the perspective that change in the policy process is likely to be gradual, and will require improved understanding by analysts and policy-makers of the heterogeneity of livelihoods and increased participation in the policy process by communities and civil-society organizations. This is most likely to come about through better understanding of existing policy processes, and where improved capacity can have the most impact. At the same time, a number of countries, particularly in Africa, have improved their public-sector management through the application of instruments for planning and economic management that are not necessarily conducive to the SL approach as presently used. The view taken here is that stable economic management combined with good public-sector practices is an important component of an enabling environment. Ways must be explored to make the two compatible and mutually supporting.
Some would argue that a more radical approach is needed (UNDP 1998b). The development of sustainable livelihood systems is a complex policy issue because of the heterogeneity of livelihoods, and because they evolve in a context of unpredictable change. Rather than simply adapt existing methods of policy analysis, which tend to be sequential and cumulative, a truly holistic approach would require the evolution of a complex adaptive management system. Such systems are beginning to be adopted in a few private-sector organizations in the United States, but there is little experience as yet in the public sector.3
There is a considerable literature on the impact of policy, particularly macro policy, on poverty and food security.4 Much of this work was initiated in response to concern about the impact of structural adjustment programmes on the poor, and it has considerable relevance to the SL approach. However, it can and should be reanalysed within a broader context.
Figure 2 shows an initial attempt to adapt an approach taken in an FAO food security manual in order to identify some of the more important linkages between policies and livelihoods, at macro, meso and micro levels. Policies, whether macroeconomic or more sectoral or structural in nature, are channelled through meso structures, which may be countrywide, such as markets 5 and legal rights, or more decentralized, through local government implementation. In some cases these two meso structures may be interrelated. At the micro level, the impact of policy on both assets and vulnerabilities may be experienced either at the individual or household level, or at the community level, which in turn affects livelihoods. The diagram abstracts from different tiers of government, but could be disaggregated to examine this in more detail. Different modes of implementation could also be included.
The SL lens focuses on assets, diversity of livelihoods and vulnerability to shocks and stresses, rather than simply on income or food-related outcomes. The linkages should not be simply one dimensional, such as those on impact on income, but should be qualitatively and quantitatively identified in terms of impact on different types of assets and also the vulnerability of the livelihood opportunities arising from those assets. For example, trade liberalization may increase the opportunities for accumulating financial assets, but also increase vulnerability to global price movements. There is still a great deal of work to be done on how decisions in various policy areas can affect not only asset accumulation but also vulnerability and adaptation.
One area that deserves broader attention is identification of the opportunities and costs for risk-proofing. This is an area that is currently being examined in the context of disaster and drought management but could be applied in a more general context.
Although, as pointed out earlier, developing appropriate policy to support and enhance sustainable livelihoods is specific to a given situation, some general desirable characteristics of a development policy can be identified. There are also specific challenges and issues that an SL-friendly development policy has to address.
Good policy, in whatever area or sector, has clearly identified objectives. In an SL context, these should be stated in terms of improving access to assets and reducing vulnerability. These need not be the only objectives of supportive development policy. Economic growth, environmental sustainability and macroeconomic stability are only a few of the other possible objectives for a development policy. These could be different elements of an overall development "vision", but that vision must be coherent and consistent and, from an SL perspective, must place SL objectives in the forefront. For the policy to have substance, the objectives must also be concrete and translatable into targets. As will be discussed in more depth later, this raises the issue of indicators of vulnerability, and how success in reaching these targets can be measured.
Sustainable livelihoods do not divide easily into sectoral pigeonholes. A given household's livelihood strategies will usually be multisectoral and interdependent. Yet most policy formulation and implementation currently undertaken is done so on a sectoral basis. Coordination and prioritization are the responsibility of planning or finance ministries. Where there is a well-constructed multisectoral policy or plan, centred on SL or anti-poverty objectives, then effective use of public expenditure techniques can provide an appropriate mechanism for achieving a coordinated policy approach. As long as implementation is along sectoral lines, however, this requires a careful exploration of the implications of the overall integrated policy for individual sectors.
Decentralization may allow for better coordination in policy implementation, and even in policy development, but much depends on the degree of autonomy and financial devolution given to the regional and district authorities.
Where policy itself is developed on a sectoral basis, this places a heavy burden on the coordination process if a coherent approach to sustainable livelihoods is to be achieved. A review of the history of developing coordinated policy for food security shows the difficulties in integrating, for example, agriculture and health policy. The sectors that have to be coordinated for an SL approach are potentially much more numerous. Note should be taken of the challenge of integrating a sustainable environmental policy with an anti-poverty approach. This was an important element in the genesis of the SL approach, and though an important aspect of individual projects, it has perhaps been sidelined in some of the more recent conceptual discussions.
Coordinated domestic policy has to be reinforced by the commitments and international policies to which both northern and southern governments agree to implement. This is important not only in order to have domestic and international policies pulling in the same direction, but also because in many important areas, such as trade policy, debt relief and conservation, SL-friendly domestic policy requires supportive international agreements.
An enabling policy framework must be clear and transparent about the role the State expects to play in enhancing sustainable livelihoods. What does the State regard as its obligations in, for example, service delivery? What kinds of safety nets, if any, does it plan to provide, and under what circumstances? What changes are planned in the legal framework that may change individual rights to property, land and access to the legal system? What role is envisaged for the private sector, and what regulation will be put in place? It is important that the State be guided by a clear anti-poverty strategy, but it is also important that the functions of the State be clearly set out. Livelihoods are often negatively affected by unpredictable State action. The rationale for State intervention and the provision of services must be clear, and implementation must be in accordance with an overall anti-poverty strategy.
The setting of realistic targets - realistic from the point of view of available public resources and also in terms of technical and administrative capacity to deliver - is an important step in moving from a commitment on paper to an active and operational policy. The challenge here, particularly for health and education services, is, within given resources, to balance quality of service with an improvement in access. This is not unique to the SL approach, but this approach places particular emphasis on improving access to assets, markets and services. One way forward may be to rely on greater prioritization of services within the community itself. This would then have to be fed back upstream into district and regional plans for the relevant ministries.
There may be areas where the trade-off between maintaining quality of services and improving access is quite limited, perhaps because the cost of maintaining access to a given quality of an asset, service or market is small. It would be useful to identify these situations where they exist.
Monitoring of policy implementation is closely linked to the setting of realistic and practical targets for policy. These targets can then be translated into measurable indicators. The area of appropriate indicators for monitoring the success of sustainable livelihoods approaches is both important and complex, and deserves more discussion than can be developed in this paper. However, certain elements can be extracted from experience so far, in particular at the micro level. (Ashley & Carney 1999). Non-income aspects of livelihoods, such as improved access and reduced vulnerability, are difficult to measure but are often better indicators of achievement than, for example, monetary income measures. At the project level, it is important to negotiate indicators with the relevant stakeholders and the poor. Some variant on this could be developed for monitoring policy progress and impact. To the extent that policy can be made more participatory in its formulation, the identification of relevant indicators, which are recognized as measures of achievement, could be part of the overall policy process. The major challenge of identifying, at national or subnational level, indicators that capture important elements of extremely diverse livelihoods is unlikely to be met completely, but it may be possible to improve on the rather aggregate measures of income often used to measure changes in poverty.
Different disciplines simplify and caricature the policy process in diverse ways (Sutton 1999). Economists often adopt simplifications that are technocratic and also linear in nature. A policy issue arises and is put on the agenda either because of pressure from agents external to government, such as aid donors or domestic lobby groups, or because it is on the government's own political agenda, which maybe expressed in some form of manifesto or from internal debate.
Once an issue is on a government's agenda, there is a process of policy analysis, where possible technical solutions to the problem are explored. These solutions arise from a bounded set, bounded by perceptions of what is politically acceptable and often by reluctance to undertake radical change. A policy is agreed on, with or without a process of consultation with stakeholders, and then implemented. Unless there is clear evidence of the policy being dysfunctional, for example through vociferous public demonstration, then the issue will be regarded as resolved.
This is clearly a simplified model (similar to that illustrated in Figure 1), the elements of which are reasonable representations of reality in some countries, but it is unlikely to describe any one country's processes accurately and may in some cases be actively misleading. However, it does have two important characteristics that are representative of many countries: it portrays the policy process as being essentially top down; and, with it, communities have an extremely limited formal role in policy process determination.
An SL-friendly policy process would indicate a much more active role for communities. Policy should reflect their own perceptions of their opportunities and their livelihood strategies, rather than relying almost entirely on those of rather remote policy-makers. Communities should play a part in setting priorities. The policy process should emphasize feedback loops rather than linear structures.
This may seem a rather utopian approach, compared with those currently found in many countries, where communities play virtually no role at all in the policy process. Until recently, good practice was where policy analysts took account of the findings of participatory rural appraisal (PRA) exercises when developing their approaches to poverty reduction. This situation is gradually changing. In Uganda, for example, a participatory poverty assessment (PPA) project has been an important input into the National Poverty Eradication Action Plan. In Namibia, both the Agricultural Policy and the National Drought Policy were extensively discussed in a series of regional workshops with stakeholder organizations, before being finalized and presented to parliament.
In the north, there is a common perception (except amongst political scientists!) that democratic process, whereby governments are elected at regular intervals, goes some way to ensuring public accountability and responsiveness, which are among the important elements of good governance. Closer examination indicates that most electoral processes give citizens little opportunity to have a direct impact on policy, hence the importance of lobbying organizations, single-issue NGOs and other forms of civil-society organizations.
Some countries in the south have introduced and funded direct participation in the formal planning process. This has taken different forms. In Uganda, the Plan for Modernisation of Agriculture (PMA), which is currently in draft form, starts the process of prioritization from the areas of action for increasing agricultural production identified by the Uganda Participatory Poverty Assessment Project. These are incorporated into the PMA, along with the priorities identified in a 1998 consultative group statement, the outcome of extensive stakeholder consultation. Poor people themselves identified increased consultation of the people, especially women, as imperative for poverty reduction, and the PMA envisages a participatory element in its implementation, for example in the development of appropriate indicators for monitoring that are location, group and farming-system specific. Uganda is clearly taking on board many of the issues arising from an SL approach. The country is still operating in a basically sectoral framework, while trying to ensure coherence among the different policy areas. The process is at an early stage and should be monitored as one possible way of creating an SL-friendly policy context.
In South Africa the new government prepared a discussion paper on agricultural policy during 1996-1998. A complex process of consultation and preparation was established, involving groups preparing drafts on key subsectoral questions. The majority of participants were from outside government, including NGOs, academics and the private sector. From these contributions, a process of synthesis was used to prepare a coherent draft, which was then discussed in a wide range of fora all over the country, which was the basis of the final version.
The use of participatory poverty assessments as a means of providing a channel of influence for the poor is being used in a number of countries. Box 1 discusses the objectives of the ongoing PPA in Pakistan.
THE PAKISTAN PARTICIPATORY POVERTY ASSESSMENT
The Pakistan PPA aims to strengthen awareness of poverty through undertaking in-depth field studies that use a variety of qualitative methods. This approach focuses on the views and experiences of the poor themselves, thereby generating a richer understanding of poverty issues. The output of these investigations is not a report but a process of change. More than a study, a PPA is a way of bringing the voice of the poor into public debates about poverty so as to promote dialogue and reflection about policy and action. Better understanding is a means to an end: that of more effective action to strengthen livelihoods, involving a wide range of actors. In fact, the purpose of the Pakistan PPA has been specified as "to give voice to the poor in the planning and implementation of policies and programmes to reduce poverty in Pakistan".
However, use of a PPA is not the only way to give people a voice in the planning process. Eritrea has adopted an approach to developing its current five-year plan for agriculture that is based on a series of participatory workshops. This started at the sub-Zhobal level, where representatives of the various kebabis, or village groupings, met to develop profiles of the agricultural sector in their areas. These were then discussed at workshops at the Zhobal, or regional level. The resulting Zhobal profiles were developed centrally into the five-year plan. This obviously begs the question of the nature of the kebabi-level committees, which represented their communities at the workshops, and the extent to which they included the positions of women and marginalized groups, a problem that exists with all participatory approaches that work through existing power structures.
Any participative process that culminates in a national policy or plan is going to face issues of aggregation. The essence of the livelihoods approach is an acknowledgement of diversity, which sits uneasily within a conventional five-year plan. Ultimately the aggregation is carried out by policy analysts and civil servants, who often end up implicitly making decisions on areas of potential conflict over priorities and access. One possible way forward is to incorporate a process of ratification of plans and policies. This however can build in an extremely time-consuming set of processes of participation in initial policy formulation and then ratification. Another possibility is for policy and plans to place more emphasis on the implementation process and incorporate aspects of participative feedback, in monitoring and assessment.
In practice, most citizens who influence policy, whether in north or south, do so by working through some collective action or civil-society organization. These can be based on activity or profession, such as unions, or on the promotion of an issue or a collection of issues, such as local and national NGOS (and, in the international arena, international NGOs) or they can be location-based committees and organizations. They can represent their members' interests in a number of ways, including lobbying, representation on public bodies and participation in consultative processes.
In many developing countries, lobbying can be seen as a difficult and risky activity. It requires resources, and specific skills, plus confidence in the robustness of the position an organization is taking. Box 2 explores some of the issues Oxfam (GB) identified when reviewing the success of its support to local partners in Kenya.
CAPACITY-BUILDING WITH CIVIL-SOCIETY PARTNERS FOR INFLUENCING POLICY
The capacity of civil-society organizations to participate in the policy process is a factor in the extent to which that process reflects and supports an improvement in livelihood opportunities for the poor and marginalized.
In a review of its support to local partners in the area of conservation, or low external-input sustainable agriculture, Oxfam in Kenya noted the lack of any impact of these activities on the policy process (Oxfam 1997). Although there had been some success at the project level, this was not being reported in a wider context, to inform policy-makers. The Kenya Agricultural Research Institute (KARI) regards NGOs as potentially influential in changing the research agenda, but Oxfam's local partners, who were generally quite small and locally based, were not exploiting this outlet. At the time of the review, they had little if any experience in lobbying or in other ways of trying to influence policy issues. Oxfam felt that one constraint was that their partners had insufficient documentation and monitoring of their project experience to present it effectively on a broader stage. The review recommended that Oxfam (GB) Kenya build the capacity of partners to engage in policy debate with research institutions and ministry officials, and, in conjunction with its local partners, improve contacts with media and then use these contacts to develop campaigns on "organic" farming.
A review of project impact carried out two years later concluded that little progress had been made on this issue (Oxfam 1999). One factor may have been a gap in knowledge, interests and perspectives between staff of local NGOs and Oxfam staff. The immediate policy issues under consideration also required the forging of a link between national policy and international markets, and therefore required collaboration between national and international partners.
Civil-society organizations also need access to policy-makers to put their points across. Unions often have easier access to ministries and politicians than issue-based organizations. In the agricultural sector, however, this access can be biased in favour of large-scale farmers, whose unions tend to be better funded and who often can employ highly skilled professional to make their case for them.
Another way that civil society organizations gain access to the policy process is by being represented on public bodies. Box 3 examines the experience in Ghana of including farmer organizations in the research process.
The scope for this kind of activity varies considerably from country to country. In many countries, quangos (for example, in the education sector, or in research) are run by nominated boards of trustees, or have advisory bodies where civil-society representation could be effectively included. This could provide a potential link between the grassroots and the policy process.
The difficulties faced by RELCs in Ghana have, in part, been attributed to a weakness of farmer-based organizations (FBOs). While some strong FBOs exist in Ghana, they are the exception. Without strong and representative FBOs, agricultural producers seem destined to continue to be underrepresented in structures such as RELCs.
STAKEHOLDER PARTICIPATION IN SETTING RESEARCH PRIORITIES
In Ghana, research-extension linkage committees (RELCs) have been formed for each of the country's five agro-ecological zones. These RELCs jointly determine research and extension priorities and plan and promote joint training sessions, field visits, workshops, field days and on-farm trials. Farmers and NGOs are supposed to participate in these sessions, with the objective of generating greater collaboration and communication among researchers, extension workers and farmers. The RELCs also aim to make research, development and transfer of technologies more responsive to the needs of the farmer, and make extensive use of PRA methods to identify key issues on the ground.
In practice, RELCs have had a limited impact. Farmers have not been properly represented, and, as a result, a recent CORAF (Le Conseil Ouest et Centre Africain pour le Recherche) report suggests that farmers' issues have not accurately been presented to the committees. The committees themselves have been dominated by crop-based research and extension staff. Women farmers and processors, in particular, have been unable to influence RELCs.
Less formally, civil-society organizations can be included in consultative processes and networks. In Namibia, for example, the Division of Rural Development in the Ministry of Agriculture, Water and Rural Development, has a rural development committee that includes members of NGOs, international agencies and civil servants, who meet on a regular basis to discuss policy and programme issues.
In Rwanda, UNDP has assisted in the establishment of a number of thematic consultation groups, which are chaired by government officials but include representation from NGOs and international and bilateral agencies. So far, some of these groups do little more than share information and identify possibilities for collaboration. Others have made a significant input in strategy development.
The hope has been expressed that decentralization will bring public resource decision-making closer to the people. As yet there is not much evidence that this is happening, but in many countries this process is at a fairly early stage. In Kenya, decentralizing has had little impact on popular involvement in policy-making, whereas in Uganda, the same type of process has had far more effect. Much depends on the degree of autonomy granted to decentralized authorities, and on the devolution of budgetary control, commensurate with responsibility. It should also be noted that decentralized government lends itself more easily to capture by vested interests, if only because the effort and finance necessary to influence policy is usually less than at the national level.
Although in principle the greater involvement of civil-society organizations should open up access to the policy process for citizens, there are a number of important preconditions. First, these organizations must themselves be open and accessible to the grassroots and, in particular, to the poor and marginalized. The organizations themselves must also feel that there are incentives to engage in policy debate, given their constrained resources and manpower. There must be a perception of potential benefit and relevance to their organizational aims.
The policy process can be seen as the interaction of a number of different types of organizations and institutions, public and private sector, and civil society, both formal and informal. Except in localized situations, individuals are unlikely to have much influence on policy but have to work through some form of organized or collective action. In a policy process that is supportive of sustainable livelihoods, organizations will tend to have certain types of attributes and structures.
It is difficult to imagine the SL approach having much influence on government organizations, whether centralized policy units or more decentralized implementation offices, unless they are:
Most of these have been elements of the numerous civil-service reform programmes that have been undertaken in a number of countries over the last decade. These have focused on improving competence and accountability through a process of organizational restructuring and revision of salary scales, among other factors. Improving accessibility and responsiveness has, in general, not been given as strong an emphasis as improving efficiency. A more efficient public sector is an important element of improved economic management and indirectly a more supportive set of structures for sustainable livelihoods. However, responsiveness and accountability to the public should also feature in an SL approach to policy processes.
The private sector has an important role to play in increasing opportunities for livelihoods. Competitiveness and efficiency are key attributes. Monopolistic and oligopolistic structures can reduce access to markets for individuals and emerging small-scale enterprises. Therefore, there is need for a regulatory presence by the State to ensure competitive behaviour and maintenance of low barriers to market entry. This also has implications for the operation of domestic financial markets and the availability of and access to financial capital.
In general it is difficult to establish appropriate incentive structures that encourage the private sector to be responsive to vulnerable groups that may have little or no economic power in the marketplace. This is particularly the case where the taxation base may be limited, thereby reducing the possibility of using tax incentives to guide private sector behaviour. This places a larger burden on regulatory systems. Access to a well-functioning legal system can be another important tool in curbing private-sector excesses, as can a well-defined set of property rights and trade regulations. Where the private sector is the preferred channel of delivery of publicly funded services, this should be controlled by a clearly defined and enforceable contract, which is properly monitored, to ensure that access to these services is equitable and that quality is of the appropriate level.
Civil-society organizations (CSOs) potentially have an important representational role to play. This may not be particularly high in the civil-society organizations' priorities, often because the returns to representing their members in wider networks may be seen, quite legitimately, as being lower than with more immediate activities. Access to resources can be a problem for all civil-society organizations, as can technical and political capacity. Where these organizations are activity based, such as with peasant farmer unions, or geographically based, as opposed to national NGOs, this can be particularly acute. At the same time, collective organization is a necessary prerequisite for any particular section of the community to have any influence on policy processes. There is a real need for capacity-building within CSOs, so that they can analyse at what level, how and over what issues they can effectively lobby or otherwise represent their members' interests in policy fora.
Another source of representation in the policy process, and indeed a locus of policy-making, are the traditional authorities in a country. The nature and importance of these authorities varies considerably, but the general principles hold; the opportunities for sustainable livelihoods for the poor will be enhanced insofar as traditional authorities are seen as legitimate, competent and accountable.
The discussion above has centred on organizations rather than institutions (the rules and regulations, both formal and informal, that determine the context within which organizations operate and the activities that they can legitimately undertake). Formal institutions are, to a large extent, the outcome of the policy process, and therefore reflect the nature and interaction of the various organizations involved in policy-making. Informal institutions, such as caste, patriarchy or religious practices, are more difficult to modify or address through the policy process, though they are often of critical importance in their impact on livelihood opportunities. In the last few decades there have been attempts to modify informal institutions through the legal process, most notably over issues such as succession to land and the rights of women and children. The process of modifying informal institutions in this way can be slow and often extremely conflictual. Its success in practice often depends on access to the legal system, supportive CSOs and resources. However, the ability to address restrictive informal institutions is an important element in achieving a policy context that is supportive of livelihoods for marginalized sections of the community. One way in which this can be achieved, albeit slowly, is to give political power to the marginalized groups. Box 4 illustrates how this is operating in Rajasthan, India.
INCLUDING WOMEN IN THE POLITICAL PROCESS IN RAJASTHAN, INDIA
Although the Indian constitution has guaranteed equal rights to women since Independence, in the more conservative regions of the country, such as Rajasthan, this has had little impact on women's position, which is circumscribed by traditions of patriarchy, feudalism, child marriage, arranged marriage and, for women of higher castes, seclusion in purdah. Eighty percent of Rajastahni women are illiterate.
In 1992, the Seventy-third Amendment to the Constitution, or Panchayat Act, was passed, which not only established elected village councils, or panchayats, throughout India, but required that one-third of the seats of these councils be reserved for women. In addition, one-third of all sarpanch, or village chief, positions had to be filled by women. These village councils deal with local civic affairs. In 1995, the first year the amendment took effect, more than 3 000 women sarpanches were elected in Rajasthan.
The extent to which women sarpanches have been able to have an impact on local affairs is mixed. However, in spite of predictions that women sarpanches would mainly be proxies for husbands or male relatives, a recent study in three of the most patriarchal states showed that two-thirds of women were exercising power on their own behalf. Almost half of those elected came from lower castes and scheduled tribes, and more than 40 percent came from households below the poverty line. Elected women members of the panchayat have provided a conduit for other women to male sarpanches, who otherwise would have been unapproachable. In some villages, improved access for women has resulted in civic funds being allocated to income-generating projects. Women have also been able to take husband and wife disputes to the village council, and in some cases have achieved more equitable outcomes than those dictated by tradition (Weaver 2000).
An SL-friendly policy process, such as the one discussed above, would allow for much greater participation in the process of setting priorities and formulating overall policy structure. However the nature of this participation and its relevance to the policy process is likely to vary according to the policy area concerned. In some cases, particularly for macro policies, the concerns of the poor have to be reconciled with the need for an economically sustainable macro-environment, and the demands of lending agencies. In other cases, countries may have signed up to international conventions that, in theory at any rate, define some of the aspects of policy choice. Figure 3 illustrates three types of policy process that could be SL friendly for different types of policy.
The three types of policy shown in Figure 3 - macro policies, public provision of services and infrastructure and institutional reforms - should be taken as illustrative rather than exclusive. For macro policies, it is unlikely that either communities or CSOs can or arguably should be involved directly in decision-making, but access to this process should be ensured for lobbying purposes. Recent international decisions on debt forgiveness have shown that lobbying can impact on macro policies, particularly at the international level. Monitoring impact on livelihoods, particularly by national bodies and through CSOs, is also an important element for ensuring that policy analysts and decision-makers consider the interests of the poor.
Once an overall fiscal framework has been established by policy-makers, there is much greater scope for participation in developing priorities and targets (this is discussed in more detail in the next section).
For institutional reforms, which define people's rights and access to assets such as land, an SL-friendly approach would centre round a highly participatory process of conventions wherein community and citizens' concerns and wishes could be expressed and explored. The formulation of the outcome of those processes into legal institutions would be undertaken at the national level, but on the basis of policy content, as identified with participatory input.
These examples are put forward as highly simplified illustrations. Actual policy formulation never exists in a vacuum, and arises out of complex political processes. However, certain types of policy are more amenable to participation by CSOs and communities, once the process is under way. Particularly where direct participation is limited, access for lobbying at various levels is critical, as is formal and informal monitoring of impact of policy on livelihoods.
The discussion so far has centred on the broad policy process. As indicated in the previous section, the delivery of publicly funded services and infrastructure is a particularly important area because participatory processes can have a direct impact on how those services are provided. This cannot be done in isolation, however, but must be undertaken within an overall macroeconomic and budgetary framework, on which national and international pressures impact.
An important part of the livelihoods agenda is to arrive at improved processes of public-sector management that link elements of policy, institutions and public-sector spending. This would combine responsiveness to locally articulated priorities with best practice in policy development and public finance.
The achievement of policy objectives is dependent on well-developed strategies. An important part of those strategies is effective management of scarce public resources. There is a growing consensus that improving the public expenditure management (PEM) system by focusing on its composition and effectiveness can increase the efficiency of government finances and donor assistance in an integrated fashion.
The systems of PEM that have been introduced in many countries around the world have four basic elements (Lawson 1999):
In order to achieve these elements, the PEM system must be transparent, predictable, accountable and comprehensive.
One particular framework that has been adopted in a number of countries as a basis for improved PEM is the medium-term expenditure framework (MTEF). This is a top-down process of determining resource availability and allocating those resources between sectors, and a bottom-up process of estimating the actual requirements of implementing policies in each sector (Oxford Policy Management 1999). Figure 4 illustrates this.
Source: OPM, 1999.
Thus there is a two-way process between the centre and line ministries, and possibly to decentralized regional authorities, which can allow for more effective costing and greater autonomy and devolution of responsibilities while maintaining overall tight control of spending priorities from the centre. The successful implementation of a MTEF is dependent on the existence of a disciplined, transparent political process of choice between public priorities. In other words, it is not a substitute for an effective policy process but a tool for translating a coherent, well-designed policy into service delivery.
The two key concepts in this are transparency and accountability. For transparency, both the macroeconomic framework and the specific targets used must be published, in a form that is accessible to the public and comprehensible, at least in general terms. The same is true for the comprehensive budget. A free media can make an important contribution to maintaining transparency and interpreting budgetary processes and outcomes for the public. Transparency is a precondition for accountability, which in itself is multilayered. The executive should be accountable both to the government and to the public for effective use of financial resources in the achievement of agreed goals. The government also should be accountable to the public for the specific objectives, priorities and targets chosen. This in turn requires effective monitoring and ex post evaluation of outcome.
In a broad sense, insofar as better PEM results in increased macroeconomic stability and more effective use of public resources, then it is supportive of the SL approach. A more predictable and economic environment increases the possibilities for livelihood enhancement.
Certain aspects of PEM and the use of the MTEF have clear affinities with the SL approach, such as:
There are also a number of entry points where linkages between SL and PEM can be incorporated:
From the perspective of donors committed to an SL approach to reducing poverty, the MTEF clarifies where donor assistance fits into the overall policy framework and the relative importance of domestic and external funding. The PEM process also gives a framework for better integration of government and donor efforts to improve the efficacy of public-sector approaches to poverty reduction.
It is also a management process that allows for the introduction of participation into a number of elements. Figure 4 identifies a bottom-up process from sectoral or line ministries to the finance ministry, but this could be extended to include decentralized authorities. The targets identified for the budgetary process could be verified in a participatory fashion. Performance reviews, in particular of the ministries providing services, could and should involve stakeholder participation, as well as a more formal accounting process.
In summary, it is not just the content of policy and the process of policy formulation that should be addressed by the SL approach. It is also the practice of policy implementation. This is an area that has been addressed in terms both of organizational reform and of financial management. There are opportunities to introduce SL concerns into the latter, and while it is unlikely that the impetus for PEM will arise from the adoption of an SL approach, it can certainly be focused in such a way as to forward SL objectives.
An SL approach to poverty reduction can change analysts' perspective at the policy level in a number of different ways:
The fundamental question, both for governments and for donors is, does an SL approach make it easier or more effective to address the issue of poverty reduction, or does it simply provide a different slant on an intractable problem? And if it does provide a practical approach to improving poverty reduction policy, in which of the areas outlined will government or donors get the greatest return for focusing their efforts? Is the policy process so interlinked with policy content that resources should be spread over both?
At present there is insufficient experience to answer any of these questions. The answers may well be different for different countries. Certain approaches may simply not be feasible in given political contexts. It will also depend on the time frame being considered. Personally, however, I think that the best possibility for achieving a sustainable improvement in livelihoods policy is to focus on increasing civil-society and stakeholder participation in the policy process. This is, however, a medium- to long-term strategy. Some would argue that in the short term, improved understanding of the impact of policy on sustainable livelihoods could give better returns through informing policy-makers.
This raises another set of issues about the present capacity and resources available to appropriate organizations to engage actively in the policy process. Do the poor and marginalized have to have a certain level of assets, in particular physical and financial assets, before they will spend time and effort to influence policy? Equally, if they do not engage directly, will there be any sustainable improvement in policy outcome leading to an increase in assets for the poor? Linked to this is the issue of identifying critical decision points for changing policy. Are there key trigger points in the policy process where stakeholder input can be most effective? One or two well-chosen case studies could be useful in answering this question..
If the rationale for increased emphasis on stakeholder participation in policy is accepted, this has implications for the focus of external assistance. More effort should be put into improving capacity and access for grassroots and community-based organizations, as well as small-scale unions. Such organizations should be encouraged and assisted in carrying out their own analysis of livelihood opportunities, and then use the results as a basis for influencing policy, through representation and lobbying.
Finally, one of the problems in writing this paper was accessing documentation on recent efforts to increase participation in the policy process. Policy content is contained in policy papers and analysis. Mission reports often contain information on the implementation of policy. However, accounts and analysis of the policy process, and in particular the development of policy, can often be found only in studies by academic political scientists, published several years after the event. If our understanding of what does and doesn't work in encouraging participation is to improve, then more effort has to go into documenting process.
Akroyd, Stephen & Alex Duncan. 1998. "The sector approach and sustainable rural livelihoods". In Sustainable rural livelihoods: what contribution can we make? Edited by Diana Carney. DFID, London.
Ashley, Caroline & Diana Carney. 1999. Sustainable livelihoods: lessons from early experience. DFID, London.
Carney, Diana, ed. 1998. "Sustainable rural livelihoods: what contribution can we make?". Paper presented at Natural Resource Adviser's Conference, July. DFID, London.
Carney, Diana et al. 2000. "Livelihood approaches compared". Rome, FAO.
Chambers, Robert & Gordon R. Conway. 1992. "Sustainable rural livelihoods: practical concepts for the 21st century". IDS Discussion Paper No. 296. February. Brighton, Institute for Development Studies.
FAO. 1998. Implications of economic policy for food security. Training materials for Agricultural Planning No. 40. Rome.
Lawson, Andrew. 1999. "The contribution of public expenditure management to government reforms". Notes on an Oxford Policy Institute Seminar. November. Oxford.
Oxfam. 1997. Conservation farming, food security and social justice. August. Nairobi and Oxford.
Oxfam. 1999. "Oxfam (GB)'s work with agricultural communities in Kenya: a review of project impact". Report for the Department for International Development, March. Nairobi and Oxford.
Oxford Policy Management. 1999. "Good practice in public expenditure management". July. Oxford. (Also on www.opml.uk).
Sutton, Rebecca. 1999. "The policy process: an overview". Overseas Development Institute Working Paper No. 118. August. London, ODI.
UNDP. 1998a. "Governance for sustainable livelihoods: operational issues". (Available at http://www.undp.org/sl/ Documents/Strategy_papers/Governance_and_SL.htm/Govern_and_SL.htm).
UNDP. 1998b. "Policy analysis and formulation for sustainable livelihoods". (Available at http://www.undp.org/sl/ Documents/Strategy_papers/Policy_Analysis_EMR/Policy_Analysis_and_Formulation.htm).
UNDP. 1998c. Policy analysis manual, with applications in Benin. New York.
Weaver, Mary Anne. 2000. "Gandhi's Daughters". New Yorker, January 10, p. 50-61.
1 The author benefited from input and discussion with a number of colleagues, in particular, Jim Gilling and Alex Duncan of Oxford Policy Management, whose contribution was funded by DFID. However, the author is solely responsible for the content of the paper.
2 Although the thrust of the argument contained in the paper is relevant to all countries, most of the examples, and the author's experience, come mainly from sub-Saharan Africa. Practice in other parts of the world may differ significantly.
3 Naresh Singh, personal communication.
4 As indicated, there is a vast literature, both theoretical and case study. For two manuals in this area, see FAO 1997 and UNDP 1998.
5 Although the actual functioning of markets may vary by district, according to factors such as infrastructure and demand levels, the overall policy on markets is usually uniform at the national level.