2. Evidence of impact

2.1 Poverty Reduction, Enhanced Resilience, and Long-Term Sustainability

In this section a general overview of types of evidence of poverty reduction, enhanced resilience, and long-term sustainability found within the cases is provided. This is followed by a discussion of the degree to which cases that demonstrated a positive impact on the rural poor had applied or incorporated sustainable livelihoods principles. The majority of case studies are described briefly within the text boxes throughout Section 2.0 while additional information can be found in the case study summaries (available in the separately printed Appendices.)

2.1.1 Poverty Reduction

Table 1 provides a general overview of how cases demonstrated improvements in the lives of the rural poor through increased income, diversification of income sources, changes in income distribution, improved basic needs and services, access to productive resources, increased agricultural yield, and changes in household food security. The scale at which these improvements were made differed greatly among cases and it was often difficult to derive numbers and types of beneficiaries from the information available. An expanded version of Table 1 can be found in the Supplemental Materials, being published separately.

Table 1. General evidence of impact on selected poverty indicators

= Evidence of an increase = Evidence of positive changes

The majority of cases reviewed demonstrated evidence of increased income associated with agricultural production while a few focused on wealth generation by initiating new non-agricultural enterprises and skills. Increases in income were the result of:

a. Increases in existing production system yields through inputs and intensification (Gambia, Ethiopia, Indonesia, Honduras, Nepal). In Gambia, the expansion and intensification of rice production transformed the situation in target villages from a rice deficit to a rice surplus. As a direct result of this, the incomes of 11,500 households increased by between 50-500%.

b. Diversification through additions of crop and livestock or other on-farm enterprises to the existing farming strategy (Ethiopia, Nepal, Bangladesh). The IRDP in Ethiopia introduced a beekeeping programme targeted at small landholders, landless men and women and the elderly, as beekeeping requires minimum labour inputs and does not take up valuable land in that hives can be placed in trees, on wasted land or even on flat rooftops.

c. Value addition to existing production (Nepal, Yemen), non-agricultural enterprises or off-farm employment. In Myanmar, the extensive labour required for physical soil conservation activities generated employment opportunities for 15,000 resource poor households over the dry season, when they would otherwise have migrated in search of work.

In most cases, access to productive resources including land, water, seeds, livestock, and trees contributed to yield and income increases, which in turn, led to improved food security and nutrition levels. In the case of Nepal, greater yields contributed to increases in household food security, such that food insecure months were reduced dramatically or eliminated during the life of the project from 9 to 0-2. However, only in a few cases were there indications of either enhanced stability/security of yields or income. In Myanmar for instance, the results of an evaluation study conducted on the sustainability of household and community livelihoods indicated that the project had made a significant contribution to enhancing the assets base of households; ensuring the sustainability of interventions and group management; and achieving a strong impact in attaining income, employment and food security increases on a sustainable basis (FAO, 2002b).

There was also evidence of improved basic needs satisfaction through increases in living conditions, nutrition, sanitation and improved access to services such as sanitation, health, education, credit, and extension services.

a. In Honduras, large-scale improvements were made in housing conditions through the installation of piped water and latrines, as well as the adoption of improved stoves. In addition, 67 community banks were established along with two cooperatives.

b. In Nepal, 15,000 women, men and children experienced improvements in health and nutrition through greater access to nutritious foods, improved domestic habits and the use of boiled water. Positive impacts were made on birthing practices, with women no longer giving birth in cowsheds at some sites and there was also a rise in the number of children attending school and receiving healthcare.

c. In Bangladesh, savings groups formed by Farmer Field School participants helped households buy and install the concrete slabs needed to improve latrines to counter the high incidence of diarrhoeal diseases in the area.

This was also confirmed through the spending choices associated with increased income reported in some cases (Yemen, Ethiopia, Nepal, Myanmar).

Very few cases contained documented evidence of changes or shifts in wealth classes, or direct impacts on the most poor. LADEP in Gambia was one of these, where greater rice yields also led to shifts in income distribution, with 15-78% of very poor project participants moving into the categories of poor and non-poor. Early and late wealth ranking exercises carried out by the DELIVERI project also showed that many farmers had moved from the ‘poor’ or ‘middle poor’ to rich categories over the life of the project.

2.1.2 Resilience and Reduction in Vulnerability

Three cases indicated increased resilience and the capacity to cope with natural or political shocks, which took the form of drought (Ethiopia,) conflict (Nepal) and climatic shock (Honduras).

In Ethiopia, the IRDP faced not only a serious drought two out of the three years in which it was implemented, but it was also confronted with outbreaks of violence and looting as the result of continuing border tensions between Ethiopia and Eritrea. The return of tens of thousands of internally displaced peoples heightened the already urgent situation in project target areas. Food for work activities linked to the rehabilitation of natural resources, road construction and other training provided a short-term solution to the immediate need for food whilst revitalising rural livelihoods in the long-term through improving the natural resource and communication bases. Soil and water conservation, reforestation and area enclosure measures introduced under the project were identified by farmers as having been critical in reducing the impact of drought on their livelihoods. Despite overall declines in terms of nutritional status and the amount of food available, project activities were successful, if not in improving the livelihoods of target community members, then in at least ensuring that the majority of target households were able to maintain their status quo against the severity of the drought (Oxfam Canada/REST, 2003).

The WIN project in certain sites in Nepal was affected by internal conflict between government forces and Maoist rebels. One village that consisted of ‘untouchable’ families for example, was highly vulnerable and insecure during the insurgency. A solution was sought based on a visit by a selected group of women from the village to a village in which drip irrigation and water tanks had been installed. The women were enthusiastic and thus guaranteed the safety of project staff, who stayed in the village for the month of work. The WIN project was seen to have helped households in other insecure areas to cope with conflict by promoting self sufficiency, strengthening groups, and community bases nurseries. Additionally, the training and team building of the WIN staff added to their willingness to continue work despite the threat this posed to their personal safety.

The Lempira Sur region of Honduras was able to withstand the ravages of El Niño and Hurricane Mitch as a direct result of project interventions. Communal natural resource recovery measures that rendered the landscape highly resistant to natural shock coupled with the subsequent introduction of new/improved production, preservation and storage technologies allowed communities in the region to maintain a grain surplus throughout the El Niño and Hurricane Mitch disasters.

2.1.3 Long-term sustainability

Franks et al. (2004) in a study of ten cases in Southern Africa, Tanzania and Uganda noted that sustainability must be considered in all of its aspects (economic, social, environmental, and institutional) in order to impact on peoples’ livelihoods. In their study, they added that economic and institutional sustainability are important for the short term while longer term consequences affecting the environment and social components must be considered.

Long-term sustainability is perhaps the most difficult to address from the case studies reviewed. This may be primarily due to the level of maturity of the various project activities rendering it impossible to comment on sustained post-project activities, institutional change, poverty reduction, or removal of exclusion or inequities among social groups within communities. However there are perhaps elements of the framework which if successfully employed can reflect a greater possibility of long-term social and institutional sustainability such as:

a. People’s empowerment - confidence, negotiating capacity, conflict resolution skills, grant-writing ability, capacity to discern useful projects based on values, education programs etc.

b. Institutional change - representation in government bodies, enhanced service provision that goes beyond an ephemeral change.

c. Enabling policies - the Bolivian water law for example.

d. Partnerships and multi-level macro-micro linkages - cohesive, multi-disciplinary teams with a strong sense of ownership and the ability to reach an expanded geographical area and multiple sectors, linkages from the community, to district, to national level meaning that successful strategies are more likely to be translated into policy.

Information that might reflect long-term sustainability in the form of environmental (mimicking ecosystems in Honduras; natural resource recovery measures in Myanmar), livelihoods (Yemen, Cambodia), institutional and community (Indonesia, Nepal, Honduras, and Yemen) sustainability was found in a limited number of cases. Financial sustainability was indicated by high rates of repayment on the loans made by community development organisations to fund income-generating and community benefit activities (Yemen, Myanmar).

2.2 Linking use of Sustainable Livelihoods Principles with Evidence of Positive Change

As shown in the previous section, evidence of positive impact on the rural poor was found in the cases under review. This section examines the degree to which selected principles, both specific and non-specific to the SLA were incorporated by the projects and attempts to draw linkages between these principles and evidence of poverty outcomes and impacts. Table 2 presents a broad overview of the nature in which the three SL-specific principles were employed by the 12 projects.

Table 2. An overview of the incorporation of SL-specific principles

2.2.1 Analyzing the vulnerability context

All cases engaged the three SL-specific principles, but to differing extents and with differing levels of success. The vulnerability context was generally characterised by food insecurity and malnutrition, a lack of disposable income, a limited asset base, the exploitation of natural resources and vulnerability to natural shocks. Efforts to address these issues mainly took the form of increasing the production of nutritious crops; promoting the sale of surplus crops to generate extra income; building human (training), social (group formation), financial (credit services), natural (planting of trees, seed and plant nurseries), and physical (treadle pumps, multi-storey trellises) capital; and the recovery of natural resource in order to decrease vulnerability to natural shocks. Boxes 6 and 7 illustrate further how these, and other principles were put into practice and how, using the cases of Honduras and Ethiopia respectively.

2.2.2 Building Assets

All cases reviewed demonstrated evidence of increases in some/all five forms of assets:

• Human asset development took the form of technical, vocational and organizational capacity-building provided to individuals, vulnerable/disadvantaged groups (poor, landless, women, female household heads, elderly, destitute), producer groups, community development associations, and local and national NGO’s and governmental institutions.

• Social assets were built through the formation, training, cohesion and capacity-building of community groups, committees, farmer groups, and local leaders. While some projects worked deliberately with pre-existing groups in the interest of cohesion and solidarity (Nepal, Gambia), others created new associations in the interest of social inclusivity (Yemen)^[9].

• Financial assets were enhanced through income generation and savings as well as access to community development funds, community banks, or other credit schemes.

• Physical assets were built through infrastructural support related to sanitation, water supply, roads, and shelter and storage facilities, and/or the provision of farming tools and other equipment (treadle pumps, multi-storey trellises, bee hives).

• Natural assets or stocks were addressed in a variety of ways, including:

  a. The incorporation of productive resources, such as the planting of mulberry and apple trees in Pakistan.

  b. Providing access to, or the reclamation of land for, agricultural production. In Gambia, backswamps and tidal swamps were reclaimed through the construction of causeways and bridges.

  c. Addressing upstream-downstream relationships through integrated watershed management, as in Bolivia.

  d. Managing lands to mimic natural ecosystems, as in Honduras. e. Promoting improved farming and land use practices, such as in Cambodia, Myanmar and Pakistan.

Box 8 presents the case of Yemen, where the building of assets through community organization, enterprise skills facilitation and access to credit and savings facilities was adopted as a key strategy in addressing rural poverty.

2.2.3 Livelihoods Focus

All cases incorporated some form of livelihoods focus. Several good examples are demonstrated in Honduras, Yemen and Ethiopia (earlier Boxes 6, 7 and 8). What was difficult to clarify from the existing documentation was the degree to which livelihood strategies were intentionally developed based on pre-existing livelihood strategies and assets analysis of the ‘beneficiary’ communities, or based on over-riding intentions of the project donor organization. Examples from Ethiopia, Pakistan, Yemen and Gambia demonstrate the divergent approaches to incorporating a livelihoods focus (Box 9).

2.2.4 Good Governance

Governance refers to the form or strength of governing systems - structure, power, effectiveness, efficiency, rights and representation and addresses inter alia exercising political power; efficiency and accessibility of service providers; honesty, efficiency, effectiveness, accountability and accessibility; human rights; property rights; and decentralization. Although not true for all cases reviewed, there were some cases in which the principles of governance and/or multi-level linkages were well illustrated, and were articulated as:

a. Strengthening customary village institutions (Gambia, Bolivia), or creating new village-level institutions (Myanmar, Yemen).

b. Building community representation in local government (Zambia, Honduras).

c. Building the capacity for participatory, multidisciplinary or collaborative approaches (Nepal, Pakistan, Honduras).

d. Enhanced responsiveness of local government to local priorities (Indonesia, Ethiopia).

e. Building the service provision capacity of government agencies (Indonesia, Nepal).

f. Influencing policy reform (Indonesia, Honduras, Bolivia, Nepal).

In most cases, there was a link from household to local government and in some cases, a link to national government. Projects that stressed aspects of governance, multi-level linkages and institutions from the outset seemed better positioned to report an enhanced responsiveness to community and farmer priorities while impacting national efforts. A brief description of the governance and multilevel aspects of the Indonesia and Honduras cases are provided in Boxes 10 and 11 respectively.

2.2.5 Social Inclusivity and Empowerment

The degree to which projects were socially inclusive and empowering varied greatly across projects. In some instances, projects specifically engaged multiple socio-economic groups including the poor and very poor (Honduras, Ethiopia, Myanmar), untouchables (Nepal, Box 11), and other marginalised groups (Yemen). In other cases, social inclusivity was expressed as including or working only/predominantly with women (Pakistan, Gambia and Bangladesh).

Projects that made a concerted effort to ensure social inclusivity were often able to facilitate the empowerment of vulnerable/marginalised groups. This was articulated as:

a. Significant changes in women’s position within the household, and access to and control of household income (Bangladesh, Box 12).

b. Significant changes in the status of other marginalised/disadvantaged groups within the community (Myanmar).

c. Enhanced problem solving by women (Nepal).

d. Promoting the rights of communities to access natural resources (Cambodia).

e. Empowering farmers to engage in extension planning (and criticism) and entrepreneurial activities promoted by women (Indonesia, Pakistan).

f. Enhanced ability to initiate and be proactive in development (Yemen).

g. Associations allowing small-scale farmers’ voices to be heard in policy debates (Gambia).

h. Reduction in dependence on or use of an intermediary patron in times of trouble (Honduras).

i. Bringing together local government and communities during project design (Ethiopia).

2.2.6 Participation

Social inclusivity and empowerment were often closely associated with the nature and quality of participatory processes put in place by a particular project. One of the aspects which contributed to strengthening the impact of the five more successful projects was precisely the strength of the participatory processes that they set in motion. Aspects of participation included:

a. Ethiopia: High levels of community participation in the problem identification, planning, implementation and monitoring and evaluation phases of the IRDP. Actions undertaken to ensure community participation in programme activities included the creation and support of local committees, interventions specifically targeting women, the landless and female-headed households, and the training of local Community Development Cadres.

b. Honduras: The completion of participatory diagnoses with groups, communities and villages within the project target area. These diagnoses were reviewed by project teams and the capacity of the project to respond was analysed. On this basis, teams negotiated a plan of work with the participating communities. At the end of the first year of implementation, and each year after that, the project carried out a process of participatory evaluation and diagnosis, the results of which were in turn reviewed by senior management and project priorities and activities altered accordingly.

c. Indonesia: The provision of training to farmers groups and Dinas Peternakan (Department of Livestock) staff to use more participatory approaches to project planning and implementation, as part of the strategy to provide more client-oriented livestock services to farmers. The Community Livestock Action Planning (CLAP) was developed as a participatory project appraisal and planning approach targeted at farmers and farmer groups in DELIVERI project villages. CLAP substantially increased DP staff understanding of livestock production issues within their districts, provided valuable background information about constraints and opportunities for livestock development, and strengthened the capacity of farmers groups to develop their own activities.

d. Myanmar: Ensuring the participation of community members through a range of community-based organisations established under the project, including Farmers Income Generating Groups, Livestock Income Generating Groups, Affinity Groups (self-help groups) and Village Forestry Groups.

e. Nepal: The formation of multi-sectoral district and national teams who were trained in participatory and gender responsive methodologies, and who carried out participatory assessments and gender action planning in local sites. Target groups included women, marginalised indigenous groups and food insecure households. These groups were trained by district staff in specialized topics, and were then supported in implementing their own plans and activities.

f. Yemen: The formation of Community Development Associations that took account of traditional power structures (by promoting the participation of local tribal leaders) whilst also ensuring the participation of poorer and marginalised community members and women. Based on a sample of 33% of all CDO members, 74% were found to be poor, representing 65% of total CDO membership. Women’s participation was somewhat lower at 37%.

2.2.7 Partnerships & Multi-level, Macro-Micro Linkages

Establishing strong partnerships proved to be a critical factor both in ensuring widespread participation and geographical coverage, and in contributing towards the longer-term sustainability of project achievements. Through some of its many partners, WIN Nepal was able to work with the non-poor (through the World Bank) and the most vulnerable groups (through GTZ). It was also able to continue working with isolated communities in areas of insurgency through its district teams. Linkages spanning the community, district and national levels were also key in facilitating the adoption of successful project strategies at the institutional level. In Gambia, 12 District Level Lowland Farmer Associations created under LADEP were linked to the National Farmers Platform and to the National Women’s Farmers Associations, facilitating the representation of local needs at the national level. As a result, attitudes towards rural development were influenced within the government, whose capacity to adopt self-help-based and demand-driven approaches was built and where the importance of combining social development with engineering works was recognised. Participatory training provided by LADEP also increased the capacity of government extension staff and transformed their way of working with rural communities.

2.3 Aspects that Challenged the Achievement of Positive Change

While there was significant evidence of the positive impacts many of the 12 projects had had on rural poverty reduction, project performance was not always favourable. Five of the projects in particular faced some/major constraints in effecting positive impacts on the rural poor (Table 3).

Table 3. Projects that faced challenges in achieving successful poverty reduction

These challenges centred around both project-related constraints (those within the power of the project to control) and wider constraints (those beyond the direct control of the project). The pattern that emerged from these challenges, discussed in more detail below, was one whereby project-related constraints appeared to be linked to the inconsistent application of some of SL-specific and non-SL-specific principles.

2.4 Project-related Constraints

2.4.1 Participation, social inclusivity, and enhancing the livelihood strategies of the poor

All projects made some attempt to mainstream participation throughout the various stages of the project cycle and to address the needs and enhance the livelihood strategies of the most vulnerable groups. A number of projects however experienced difficulties in this due to design weaknesses.

Local stakeholders were not consulted prior to the design of CARE’s SHABGE project in Bangladesh for example, and difficulties were experienced in encouraging women’s participation due to strict socio-religious codes limiting women’s mobility and presence in the public sphere. In one district, it took facilitators four months to satisfy the minimum participation requirement of 20 women and only after extensive negotiation with husbands, elites and local politicians to explain the project’s goal and strategy. Women’s lack of involvement in project design also had implications for the relevance of these activities to their needs. As one PNGO leader put it, ‘only about 60-70% of women members are now participating in FFS sessions. The others have stopped participating because they don’t see the project as a way of reducing their poverty’ (cited in Wilson & Hussain, 2002: 19).

The Participatory Upland Conservation Development programme in Bolivia experienced similar problems. While the PUCD programme made a number of advances related to raising awareness, enhancing income of some groups, good governance and impacting national policies, some difficulty was experienced in enhancing the living conditions of the poor and landless, including women. The PUCD found that participatory processes were not sufficient on their own to ensure the equitable participation of socially marginalized groups. By 1997, only 14% of the 202 households participating in technical training ‘hard’ project activities^[15] belonged to landless households, or to those owning less that 3 ha of land. Similarly, by 1999, women constituted just 15% of all project participants, with the highest percentage concentrated in ‘soft’ activities^[16].

The reasons for these weaknesses were twofold. Firstly, greater attention could have been paid to the nature of the livelihoods strategies of poor and marginalised groups, and the fact that the limited range of assets open to them often prevented them from qualifying for project assistance. Secondly, greater attention by project implementers could have been addressed to meeting the demands women had expressed during participatory planning sessions. This was reflected in women’s poor participation.

2.4.2 Issues surrounding the disaggregation of project interventions

Related to the lack of consistency in targeting the most vulnerable groups was the tendency to categorise ‘the poor’ as a homogenous category. The SHABGE project illustrates this well. While it was aimed at ‘poor and marginalised men and women farmers’, a lack of systematic selection criteria meant that project participants were selected somewhat arbitrarily by Field Trainers (FTs). Access to land was used as a key indicator of poverty, and while some FTs were satisfied if two-thirds of the households in their FFS had less than 25 decimals (1013m²), others selected only those with 10 decimals (405m²) or less. Other staff, having been instructed to focus on ‘the poorest of the poor’ had selected landless families (Bartlett, 2002).

Because of these differences in levels of land ownership/access amongst FFS participants, benefits gained from homestead gardening interventions also differed. Those with greater access to land gained greater benefits from homestead gardening activities. Conversely, those with limited access to land reported that their homestead spaces were so small that little or no income was generated from selling vegetables. Such participants were also unable to afford inputs such as seeds, seedlings, fencing or irrigation equipment in order to make the limited land they had available more productive.

2.4.3 Issues surrounding empowerment

Measures to build human and social capital had generally contributed to the empowerment of beneficiary communities, as in the case of Yemen. With some projects however, the nature and execution of project interventions limited the extent to which beneficiaries could be ‘empowered’. In Bangladesh for example, FFS participants complained that they were unable to contribute to processes designed to keep track of and illustrate changes that were occurring on study plots because they were illiterate. Calls were made for basic literary training to be held prior to the implementation of such components. Instead, the learning process was simplified in order to make it more user-friendly, although this was done to such an extent as to render it almost meaningless (Bartlett, 2002).

Concerns were also raised as to the quality of training women received from the FFS. A trade-off made between the depth and the breath of this training meant that participants had become ‘adopters’ of, and not experts in new varieties, practices and technologies (Bartlett, 2002), able to recognize, but not to understand the benefits arising from them. A tendency by FTs to view the FFS concept as prescriptive rather than flexible may help to explain this. Field staff were hesitant to adapt FFS topics. They also tended to be unclear as to the study plot objectives and were often the ones to decide the nature of FFS activities, giving participants little ownership of the learning process.

2.4.4 Issues concerning holistic interventions, increased resilience and ability to withstand shock

Most projects were not based on holistic diagnoses. The LADEP Gambia project, in attempting to the increase levels of food security and raise the incomes of impoverished household through the promotion of monoculture rice production in lowland areas, may have increased these households’ vulnerability to other, different shocks. An increase in the number of rice farmers by up to 200% in some areas has resulted in human and financial resources being concentrated into rice production at the expense of other (upland) crops.

Some villages witnessed a reduction in (1) the production of crops such as groundnuts, where men (traditionally upland farmers) had chosen to switch from groundnuts to rice because of higher returns, and (2) vegetable production on homestead gardens, where labour requirements for dike construction reduced the amount of time women were able to spend on homestead gardening and where they considered rice more profitable anyway.

These changes have potentially negative implications in terms of both nutritional levels (lack of diversity in the diet)^[17] and increased vulnerability to natural shock (pests, drought). In addition, an increase in the amount of standing water behind dikes for longer times than previously had anecdotally contributed to an increased incidence of malaria in the 11 project sites, increasing households’ vulnerability to health-related shocks^[18].

2.4.5 Issues surrounding engaging dynamism and flexibility

In Indonesia, the DELIVERI project worked in collaboration with government agencies in order to develop responsive and quality service delivery related to livestock. Peter Bazely (1999, http://www.livelihoods.org/static/pbazeleynn119.html) noted that difficulties arose when the project parties could not easily conceptualize a project that was more "non-physical" in nature, and were not as willing to engage in this. Additionally, contractors found it difficult to work with such a flexible effort that focused on transforming structures versus delivering tangible products.

2.4.6 Issues surrounding good governance and institutions, and macro-micro linkages

Despite weaknesses in its approach, SHABGE made important contributions to the empowerment of women. Participants reported tangible improvements to their status within the household and community in the form of increased incomes, a greater role in household decision-making and greater mobility, and intangible impacts such as a greater sense of self-confidence and self-worth. Although SHABGE had great potential to support these women to realise changes in their status, it did not engage district and national government representatives, undermining the institutionalisation of these achievements.

2.5 Wider Constraints

2.5.1 Issues surrounding partnerships

Examples such as WIN illustrate well the idea that working in partnership is an effective means of addressing the multi-dimensional nature of rural poverty. While the IHFSAN project in Zambia (Box 14) attempted to work in a similar manner in order to address multi-sectoral concerns, institutional and policy changes underway when the project was initiated were key in explaining the weakness of partnerships established under the project, and the subsequent lack of significant achievements made despite the numerous development initiatives that were undertaken (FAO, 2004).

The case studies reviewed have demonstrated principles and aspects that have worked well and those that have worked less well. Cases that have been successful in improving the lives of the rural poor were rarely without constraints that required attention. A number of projects for which constraints were evident chose to embrace the learning and have indicated that they were addressed or would be addressed in the upcoming phases (Gambia).