Independent Evaluation - Home > Search

         Generating Income for Afghan Refugees in Pakistan

Goals and implementation


By 1980, more than three million Afghan refugees, accompanied by as many head of livestock, had entered Pakistan to escape the war in their homeland. Most of them settled outside cities or in rural areas near the border, where they sometimes outnumbered Pakistanis. The refugees' daily requirements of fodder, fuelwood, water, and other supplies caused damage to already stressed rangeland and forest areas, and they overloaded rural infrastructure.

After a massive international relief effort met the most urgent humanitarian needs, the United Nations High Commissioner for Refugees (UNHCR) turned its attention to the refugees' needs for supplemental income and to repairing the damage to natural resources and infrastructure. A donor trust fund, administered by the World Bank, has supported the implementation of about 295 subprojects since 1984. Financing was provided by Britain, Canada, the European Union, Finland, the Netherlands, Norway, Sweden, Switzerland, and the United States. Germany, the largest single donor, funded IGPRA directly. Almost two-thirds of project expenditures have been in North West Frontier Province (NWFP), one-third in Balochistan, and about 5 percent in Punjab. (See table.)


The main goals of IGPRA were to:
    --create jobs and income, mainly for Afghan refugees but also for the local poor, through labor-intensive rural projects;

    --repair some of the physical damage the refugees and their livestock had caused to infrastructure and the environment; and

    --create lasting assets for the host country.

The subprojects were designed to be cost-effective, short term, environmentally and technically sound, and labor intensive (with labor accounting for at least 60 percent of investment costs). Subprojects, involving reforestation, catchment management, irrigation and drainage systems repairs, flood protection, river training, and road improvements, would obtain at least half their labor force from refugee populations. Some forestry and road subprojects, and all the flood protection and river training subprojects, tackled problems that existed before the arrival of Afghan refugees.

The Bank decided to use existing provincial government agencies, rather than create new ones for the project. Implementation procedures were the same as those used by the provincial governments in similar public programs. Subprojects were carried out by provincial line departments, with coordination by a small project unit in the federal government's States and Frontier Regions Division, and with assistance from short-term, internationally recruited technical experts.

Evaluation methods

As part of the OED evaluation, a team of Pakistani consultants conducted a beneficiary social impact study. The evaluation made use of existing sources of information, validated by extensive site visits and field interviews with refugees and local project staff. Tracking the extent of damage repairs was difficult because there were no baseline surveys, and most of the areas were already under stress from over-use by people and animals before the refugees arrived.


The achievements of IGPRA are impressive in the substantial range of assets created, the relatively problem-free implementation by local agencies, and the impact on refugees' employment and incomes. IGPRA exceeded its employment creation objectives and in 12 years generated 24 million workdays of labor, of which three-fourths went to Afghan refugees. IGPRA provided an estimated 11 percent of the employment needs of the available refugee labor force and improved the skills of Afghan workers. IGPRA II and III provided formal training in forestry management and, for Afghan women, instruction in producing tree seedlings and conserving energy at home. Given the enormous number of refugees involved, a much larger program would have been needed to provide income for all employable Afghans.

IGPRA made a major contribution to reversing and repairing damage to infrastructure. Examples include irrigation and road works in the suburbs of Peshawar, roads near refugee camps, reforestation, and major roads that serve as supply routes for refugees. The project successfully created durable assets for local populations and the host country. Subprojects restored ground and tree cover, and reduced soil erosion, surface runoff, and flood hazards.


Compared with similar rural programs in other countries, the subprojects were cost-effective. Field inquiries verified that, with few exceptions, subprojects made good use of resources. Where unit cost comparisons were available (such as for roads), they were in line with the costs of similar works in comparable countries, and with experience elsewhere in similar types of cash-for-work programs. Economic rate of return estimates made when the projects were approved could not be verified or reworked because no supporting data were available.


Employment, skills

The direct impact of IGPRA employment on individual refugee families, though significant, was temporary. Few refugees were able to work on IGPRA for more than a few months at a time because of the seasonal and short-term nature of the subprojects.

The full employment impact of IGPRA was higher than indicated by the number of labor days created directly by the projects because refugees used the skills acquired while working on subprojects to find work elsewhere in Pakistan. Formal training provided under IGPRA had a limited impact, but on-the-job training gave workers skills in reforestation, irrigation and drainage systems, road improvements, and flood protection. Some refugees formed labor pools with continuing ties to individual contractors, carving out a niche for themselves as sources of labor. Skills developed by refugees will also be useful in reconstruction programs in Afghanistan.


Because Afghan custom dictates that women be generally segregated from men, women did not participate in the implementation of most IGPRA subprojects. Although most projects were not designed to provide employment specifically to women, a few Afghan women were employed in seedling production at home. More could have been done to employ women in activities compatible with Afghan customs.

Women were indirect beneficiaries of the subprojects and obtained both positive and negative results from them. Women generally preferred their men to work on IGPRA because the proximity of subprojects enhanced women's sense of security—since men returned home in the evenings in most cases. But where land near the refugee camps was closed for forestry and conservation subprojects, women's burdens increased because they were forced to travel farther to gather fuelwood and fodder or had to buy these items.

Natural resources, assets

Given the limited scale of IGPRA, compared with the enormous impact of three million refugees and their herds on natural resources, the remaining problems are substantial. In some areas, damage repair is impractical as long as refugees and livestock remain in significant numbers.

However, the net environmental impact of IGPRA was positive. Flood protection works reduced losses of productive land and village buildings. New areas of forest were planted, reducing pressure on remaining forests, and conservation works protected catchments, helping to reduce soil erosion and flood hazards. Some local landowners benefited from IGPRA training in forestry management.

Subprojects in NWFP and Punjab have had a clearly beneficial impact on vegetation and in establishing plantations in severely degraded lands. Environmental impact assessments in Balochistan identified beneficial impacts from the regeneration of rangelands and tree cover, as well as from soil and water conservation measures. An unexpected benefit in Balochistan was an increase in wildlife in some subproject areas.

New and improved rural roads increased access to isolated areas, improving market opportunities, but their environmental impact is not yet clear. It is also uncertain what effect the conversion of traditional grazing lands to protected plantations will have on the environment. The plantations presently provide grass for stall-fed animals, but fodder yields will diminish as the plantations develop a closed canopy, and alternative fodder sources will need to be found in the next four to ten years.

Local populations and equity

IGPRA often increased local inequities. Although one IGPRA objective was to benefit local communities, some river training subprojects benefited only a handful of landowners and provided substantial benefits to well-to-do farmers. The rural poor were ignored and in some locations suffered because lands that had previously been accessible and communal were closed. With the closures, only those with "senior" property rights were allowed to use the lands, while those who had previously held "junior" rights to grazing and fuelwood collection were excluded. Most flood protection works benefited private fruit orchards. For refugees, IGPRA wages supplemented the subsistence needs unmet by relief supplies. Many local Pakistanis, however, considered IGPRA wages to be too low to support a family (except for the very poor).


Most infrastructure built by IGPRA is likely to be sustained. Problems are evident in 15 of the 60 infrastructure projects inspected in the field, in particular where assets (such as heavily used roads) are not being maintained. IGPRA III included an extension component aimed at organizing local communities to take over the operation and maintenance of completed subprojects, on the assumption that IGPRA would not be extended into a fourth phase. Some of the assets created by IGPRA will depend on the government for continued maintenance.

Judging the sustainability of forestry and conservation subprojects is difficult, since climate, land tenure, and social factors are all involved. A number of forestry subprojects on state lands in Balochistan are clearly not sustainable and should not have been pursued, given existing technical problems. Experience from other projects in NWFP and Balochistan suggests that sustainability is more likely if local communities are more involved in design and implementation. This was not always the case with IGPRA.


IGPRA's approach might be duplicated in other countries, but only after careful analysis of their specific circumstances. Some of the conditions that made IGPRA successful in Pakistan-especially the cultural affinities between refugees and hosts-may not exist elsewhere, and even within IGPRA there were significant variations from one refugee area to the other. A useful strategy for similar programs elsewhere may be to select subprojects that promote skills needed in reconstruction after repatriation.

A clear understanding of cultural affinity, or the lack of it, in each situation is important. Cultural affinities between the refugees and the host population contributed greatly to the success of IGPRA and are a potential source of strength in planning and executing activities in other refugee situations.

Using existing local bodies is preferable to creating new and special institutions for project implementation.

Project selection in practice was dominated by the goal of creating jobs for refugees. IGPRA could have benefited more from the application of well-defined environmental, developmental, and equity criteria for project selection. Locating subprojects near refugee camps was not essential for meeting employment goals and may have restricted the selection of subprojects. Instead of treating subprojects as discrete interventions, the project could have selected developmental priorities for each geographic region. Subprojects could then have been clustered around those priorities.

A link between the production of resources and the needs of refugees may help conserve natural resources. Environmental damage by refugees being a major concern, some of the subprojects should have been designed to cater better to refugee demand for fuelwood and fodder, which led to much of the damage.

A proper understanding of local tenure rights and an understanding of the potential impact of subproject intervention on different segments of the local population should be a part of project selection and preparation. Existing inequities were exacerbated by IGPRA because land tenure issues were overlooked. More attention in the planning phase to the implications of projects could avoid adverse impacts on women in the future. Targeting families most in need of jobs could also have improved the project's social and equity impact.

Sustainability could have been increased substantially if local communities had been involved in the selection, design, and implementation of the projects, thus increasing their sense of ownership of the projects and the assets created, and if rural organizations had been successfully organized.

© 2012 The World Bank Group, All Rights Reserved. Terms and Conditions