33. Project design. The project design was complex and perhaps overly ambitious. It
had many components and activities (e.g. gender, infrastructure, community
development, project management) and stakeholders (Ministry of Rural
Development; Ministry of Water Resources and Meteorology; Ministry of Women‟s
Affairs; Ministry of Agriculture, Forestry and Fisheries; IFAD; GTZ; United Nations
Development Programme; and AusAID), and built on two quite different GTZ
projects. In addition, although the commune council elections were foreseen at
appraisal, the advent of these councils dramatically changed the delivery, support
structure and mechanism at provincial and district level. For example, provincial
departments initially played a major role in the identification, planning and
implementation of project activities, in particular infrastructure activities. This
situation changed as the commune councils became more active and took over
responsibility for the development of their communes.21 At the time of project
design the institutional changes were already on the horizon but were not
confirmed so it might be considered to be risky to design a project delivery
mechanism (decentralized coordination mechanism) based on a still uncertain
political situation. However, CBRDP‟s alignment with the National Seila Programme,
which was designed by the Government to coordinate external assistance and
national development efforts with regard to decentralization, facilitated the
implementation process.
34. Initially gender was not mainstreamed in a coherent manner throughout the
project (it was only clearly identified as a sub-component of community
development). However, after the MTR, gender became a separate objective with
its own, albeit limited, resources. Gender training activities for staff in other
components were limited, and consequently gender considerations were not fully
integrated into all project activities, and were not explicitly monitored with, for
example, systematic assessments of capacity as foreseen in the revised logframe
of the MTR.
35. CBRDP had six outputs with 44 indicators that contributed to the project‟s
objective. The PCR notes that this excessive number of indicators contributed to
insufficient data collection, and IOE concurs with this assessment.
36. There were some design faults, such as the selection of some of the irrigation
systems to be rehabilitated (e.g. some irrigation systems were constructed on
marginally irrigable land and/or were poorly designed22). This lead to supressed
yields, inadequate dry season command areas, low water productivity, and reduced
farmer income which in turn, suppressed farm profitability and resulted in farmer
dissatisfaction with services, and unwillingness and incapacity to pay irrigation
service fees (ISFs).
37. Relevance to the needs of the rural poor. The large majority of rural
households in the two target provinces, Kampong Thom and Kampot, are farmers
with farming systems dominated by rice. In 2001, low rice harvest yields did not
meet subsistence needs, and diversification of income opportunities was therefore
required. There was also limited basic infrastructure such as roads, water supply
and irrigation, as well as very limited capacities to upgrade them.
38. Satisfaction rates expressed by villagers in beneficiary impact assessments, as well
as provision of labour and time in LTCs and water users committees, suggest that
the activities were relevant to their needs.
39. Targeting. The project design focused on crop demonstration and extension
activities that helped poor rural households who had access to some land, while the
landless were likely to be the poorest group were missed by this. The irrigation
component rehabilitated existing irrigation schemes, thereby providing water to households who may already have had some access to it. Similarly, project training
and promotion of improved methods for raising cattle and buffaloes were of more
interest to the households that already owned these animals. However, the femaleheaded
households in the project areas (21 per cent of direct beneficiaries) tended
to have, on average, less land, fewer and lower total value of assets; in other
words, these were among the poorer households. Therefore, the direct
beneficiaries of the project were a mix of households whose depth of poverty
ranged from the poorest of the poor to the less poor, with the latter category
benefitting most and the former category benefitting the least from the agricultural
component.