Interim Evaluation

The Root and Tuber Improvement Programme, (RTIP) became effective in January 1999 and is scheduled to close in December 2004. IFAD and the Government of Ghana are considering a second phase for this project. This evaluation of RTIP was undertaken at the request of the IFAD Africa I Division as a precondition for the formulation of any second phase project. It took place over the period May – November 2003. The evaluation team conducted fieldwork in late July and early August, including visits to 11 districts and meetings with stakeholders in areas of RTIP activities as well as in Accra, Kumasi and Rome. Further, it drew upon a Beneficiary Assessment Study (BAS), commissioned by RTIP¹.

The evaluation team used the IFAD Office of Evaluation Methodological Framework for Project Evaluation that emphasises consultation and participation of partners with the expectation that the knowledge they acquire will contribute to their improved performance in future. For this, the Core Learning Partnership¹ was established to guide the work. This group reviewed and cleared an Approach Paper that outlined the evaluation methodology, key questions, and calendar.² It held briefings and de-briefings with the evaluation team prior to and following fieldwork. It will review the draft evaluation report and meet to conclude an agreement among partners, known as the Agreement at Completion Point (ACP).

The project rationale justified an RTIP focus on root and tuber crops because of their importance in household food security, because they are grown by the poorest segment of the population, because their development would help diversify agricultural sector development, and because supporting them would channel more agricultural sector resources to the smallholders who produce, process and market them. RTIP was designed to be a nationwide project, referred to as a programme that targeted some 720 000 resource poor farmers, with priority to be given to women.

The overall objective of RTIP was to enhance food security and improve incomes of resource poor farmers. Originally this was to be done through five components: i) multiplication and distribution of improved material ii) integrated pest management; iii) adaptive research; iv) community support and mobilisation, and, v) programme management and coordination. A sixth component for post-production and marketing was added subsequent to a project start-up workshop in August 1998. The Ministry of Food and Agriculture, Crop Services Division, was assigned responsibility for RTIP implementation and the World Bank was appointed as the cooperating institution for project supervision and implementation support. Numerous research institutions, universities, non-governmental organisations and other agencies were assigned responsibilities for implementation. Liaisons with the International Institute for Tropical Agriculture (IITA) and with other IFAD projects in Ghana were also important to project implementation.

One major policy change, the President's Special Initiative on Agribusiness/Cassava, (PSI) affected the RTIP implementation environment. An agreement was concluded between PSI and RTIP whereby RTIP was to supply improved planting material to cassava farmers providing inputs to PSI starch factories. It raised awareness about cassava and increased demand for RTIP planting materials to the benefit of RTIP. However, the considerable efforts that RTIP made to respect its PSI agreement detracted from efforts to realise other elements of its programme and PSI's different approach sometimes confused farmers.

The evaluation found that the rationale for RTIP was very relevant for rural poverty reduction in the Ghanaian context at the time that RTIP was designed, as were RTIP objectives. The components that made up RTIP were also relevant to achieving those objectives. However, the original omission of post-production and marketing activities from the original programme was a major flaw in its design. The component subsequently added to cover this area was inadequate, as was the priority afforded by the PCO. This has been the most important factor that has kept RTIP from reaching its overall objectives so far.

Looking at RTIP results, the planting materials multiplication system established an efficient three tier system for multiplication and distribution of four improved cassava varieties in 50 districts between 1999 and 2002. The recorded number of farmers who have accessed new materials is about 105 000. If those farmers that have accessed improved varieties but have not been recorded were to be included, the number would be considerably higher. For sweet potatoes, two improved varieties have been multiplied and distributed to a total of some 14 500 farmers. Attention to yam and cocoyam by RTIP has been very limited, with no significant achievements to date in the multiplication and distribution of improved varieties of these crops. This was largely due to the fact that there were no improved varieties available to multiply and distribute. Actions that were undertaken included promotion and multiplication of two highly valued local yam cultivars as well as a campaign for propagation of coco-yam late in the programme.

The component for adaptive research served as a vehicle for undertaking more than 60 research projects submitted by Ghanaian researchers in more than a dozen different government and academic institutions on agronomic areas as provided for in the RTIP appraisal report. During the course of implementation RTIP added investigation into integrated pest management, post production and marketing to address newly arising issues. Among this component's results are the release of five new cassava varieties, upcoming releases of sweet potato, yam, cocoyam and more cassava varieties currently in the pipeline, recommended practices for control of Imperata cylindrica, control of tuber rot and options for maintaining soil fertility. Overall, the activities have contributed significantly to the knowledge base in Ghana on root and tuber crops. However, the component limitations included its approach to determination of research priorities, inadequate consideration of socio-economic factors in the selection and evaluation of varieties and practices to be pursued for development, limited dissemination of research results, weak links with extension, and minimal integration with the workings of the PCO and other components.

In the IPM component results have been below target due to delays in construction of two of the three planned insectaries. However, predators furnished by the RTIP appear to have led to the successful control of the Green Mite Mononychelus tanajoa in seven of the ten regions of Ghana. Efforts to control Larger Grain Borer were made on a limited scale, with correspondingly limited results to date. Seventeen Farmer Field Schools, established in fifteen districts, served as vehicles for IPM and basic root crop cultivation techniques. The evaluation team considers the FFS to have been successful, but arguably inadequate in number and too resource intensive, reaching a total of only about 600-700 farmers in 20% of the 76 districts currently in RTIP. Further, FFSs had weak links with the adaptive research component resulting in missed opportunities to the detriment of both researchers and beneficiaries.

In the Community Support and Mobilisation component, RTIP has formed some 9 800 groups, about 96% of the groups were for production, 3% processing and less than 1% marketing. By and large groups were formed for specific temporary purposes. The evaluation team considers that groups formed have served as effective tools for the distribution of planting material and transmission of knowledge on improved cultivation practices. However, according to the BAS only 3% of respondents liked the group formation efforts and an estimated 80% of groups formed were either non-functional or disintegrated at the time of the study in July 2003. Processing and marketing groups, often based on pre-existing relationships among members, were more cohesive. Other efforts funded under this component included upgrading of government and NGO staff skills in outreach to beneficiaries, training of beneficiaries in improved crop utilisation, strengthening links to private sector processors, and an information, education and communication campaign. These were implemented roughly as anticipated, but suffered from a lack of a vision or overall strategy pulling together these disparate activities in a unified effort with a clearly defined purpose.

Under the post production and marketing component, RTIP collected information related to storage, processing and utilisation of roots and tubers, but stopped short of fully exploiting that information. Similarly, a study of marketing issues was undertaken. Improved cassava graters, stoves for roasting and screw presses were developed. Training modules were developed in cassava processing and in use of sweet potatoes. They were delivered to more than 1 500 people including MOFA staff, small-scale processors, bakers, and farmers. However, the PCO did not secure the staff and technical expertise to give this component the due importance. Cooperation between RTIP and the IFAD-funded Village Infrastructure Project in this field never materialised, despite being formalised in a Project Working Agreement in November 1998 as a condition of loan effectiveness. The lack of attention by MOFA and RTIP's PCO to the resolution of these problems is surprising given the seriousness of post harvest and marketing issues facing the farmers.

Results and performance in the Programme Management component were well above average, rising to the challenge of the design of this ambitious nationwide project. The PCO succeeded in creating working relationships with numerous different organisations and institutions throughout the country. It made substantial achievements in plant multiplication and distribution, integrated pest management and adaptive research. By and large, the PCO administration of the implementation of physical and financial activities, including disbursement, accounting and reporting, has also been good. The PCO showed weaknesses in two respects. Firstly, it tended to focus more on the technical and scientific sides of what was needed to achieve its overall goal, shying away from fully dedicating itself to some of the harder to solve, but essential issues on the economic and social side of the investment. Secondly, RTIP management appears to have allowed its energies to become fully absorbed in the logistics and practical details of achieving the physical and financial targets. Given the size and complexity of the task this is understandable. Nonetheless, it took insufficient time to consider to whether RTIP's implementation was leading to the desired longer term outcomes and impact.

RTIP, to its credit, has been reasonably effective in reaching the specific objectives of its individual components, even if there have been shortcomings and weaknesses in each component as alluded to in the descriptions of component results and performance above. It has concentrated most of its efforts on and been most effective in the planting material multiplication and distribution component, while it has been least effective in the added-on component for post-production and marketing.

It is still somewhat early to judge how effective RTIP in its entirety will be in achieving at completion its overall goals of enhanced food security and improved incomes of resource poor farmers. The RTIP log frame targets identified three indicators for achieving its goals: 720 000 beneficiaries reached; calorie consumption during lean season in beneficiary households increased by 20%; incomes in beneficiary households increased by 15%. RTIP records show it having reached 120 000 households to date. Nonetheless, the evaluation team considers that it may well reach close to 720 000 households by the end of the project. RTIP does not have records on income or calorie consumption that would allow similar assessments of whether it will be effective in reaching its targets in those areas.

Information collected by the evaluation team and drawn from the BAS indicate that there are very likely to have been income increases above 15% for the households of the more than 2 000 farmers who have participated in RTIP as secondary multipliers. In addition, the 14 500 sweet potato growers who adopted RTIP distributed varieties had output increases and faced ready markets are also very likely to have realized such income gains. However, it appears that the majority of ordinary farmers who adopted improved cassava varieties will have had little if any income increases.

Based on RTIP data the evaluation team considers that ordinary cassava farmers have achieved yield increases of up to 40%. For this reason, despite the absence of data on calorie consumption, the team considers that the food security of participating households has improved. Yet increased yields have not translated directly into increased incomes for ordinary farmers for several reasons. Firstly, revenue from cassava occupies a relatively limited share of total household revenue. Furthermore, there has been local inflation of 65% during the period January 2001 to May 2003, increased production costs associated with RTIP-introduced practices and varieties, relatively lower prices for the RTIP varieties, and a decline in cassava prices generally – most likely due to the overall increase in output levels in local markets as a result of RTIP and PSI³.

Notwithstanding the fact that RTIP's ultimate impact on the household income of resource poor cassava farmers remains to be seen, there are some indications that the investment will prove to have been relatively efficient in terms of its overall benefits to Ghana and the Ghanaian economy. Roughly speaking, the evaluation team estimates that there have been food security benefits for some 80 000 producer households (out of about 120 000 reached) and income benefits for some 16 000 households. Other, observable economy-wide benefits to-date from the RTIP investment of USD ten million, (about USD 14 per beneficiary household, based on the total number of beneficiaries targeted) include improved work processes and systems related to support of these crops by MOFA and other agricultural sector institutions. There have been lower consumer prices for root and tuber products. Further, RTIP has fostered a nascent potential for cassava exports, for low cost cassava supplies to agro-processing industries and for decreased dependency on imported wheat flour. The utilization of project facilities and services is high. On the whole the standards of RTIP services have been high despite some variation by component. Actual costs or expenditures have corresponded, for the most part, to appraisal estimates and there have been no significant implementation delays.
18. The evaluation assessed the rural poverty impact of RTIP in the six standard domains of impact set by IFAD: financial and physical assets, human assets, social capital, food security, environment and institutions and policies. The evaluation team rated RTIP overall impact on beneficiaries as modest.

So far the project has had its most widespread impact in the domain of food security. It is considered very likely that food security improvements will spread as RTIP reaches more households and that the changes in food security will be sustained given RTIP's impact on the farming technology and practices.

RTIP also made a positive impact in the domain of human assets in terms of access to information. The information, education and communication campaign of RTIP, including a national media campaign, training activities, and Farmer Field Schools have all contributed to achievement of this impact by making knowledge of the production, uses, processing and marketing of root and tuber crops more widespread among resource poor farmers. By way of doing this RTIP also, necessarily, deepened and widened available information on these crops among the scientific community, government staff and the public at large. In the same vein, RTIP had a positive, albeit limited, impact on the public sector institutions associated with its implementation and the services they provided.

As noted above, RTIP has not had the impact on household income levels on the scale that it aspired to. Nor has RTIP made a significant impact on social capital and empowerment of resource poor households that it might have had were more attention to have been given to these matters. The evaluation team found no evidence that RTIP has had any impact on policies or regulatory frameworks that affect the rural poor. Whereas, it will very likely have had a negative impact on the environment in so far as increased and intensified cassava production worsens soil nutrition status.

An estimated 39% of recorded RTIP beneficiaries were women, however RTIP efforts to ensure equitable impact, by gender, were limited. This was a weakness in both the design and the implementation of the project. Notwithstanding the fact that women were to be considered priority members of the RTIP target group, according to the Appraisal Report, the project did not explicitly monitor its impact on women, nor did it systematically disaggregate data by gender in its records or analyzes. RTIP did not adequately compensate for the missing collaboration with the IFAD-financed VIP project that was to involve women through its activities in processing of root and tuber crops. The project overlooked the opportunity it had to take into consideration gender specific needs and priorities in the implementation of RTIP research and extension components. Likewise, it neglected to actively promote the participation of women in the secondary and tertiary multiplication of improved planting materials.

The evaluation team has rated as "highly likely" the sustainability of RTIP impact in those areas where change was farthest reaching, i.e. food security (including farming technology, agricultural production and lower frequency of food shortages) and access to information. Whereas, it is considered unlikely that RTIP's modest impact on social capital and on services provided by public sector institutions will be sustained without further support.

The design of RTIP, nation-wide in scope with a focus on one commodity, was itself an innovation at a time when the lion's share of IFAD financing was going to area-based rural development projects, irrigation and rural finance. Further, the RTIP design responds to the identification of roots and tubers as "poor man's" commodities. The assumption that by supporting the development of these crops RTIP would benefit the poor people who grew them was innovative as an approach to poverty reduction. Yet it proved to be only partly true. The innovative choice to focus on a commodity was useful as far as it went, but it had limited impact on the levels of poverty of its producers in so far as it focussed primarily on production and almost exclusively on production of cassava. Insufficient attention was given to the characteristics of new varieties in the light of the role of the selected crop in the economy of a resource poor household.

As a nation-wide project RTIP cannot be up-scaled, however it does offer some useful lessons with respect to replicability. In particular, RTIP's three-tier planting materials multiplication and distribution system proved to be a successful one that merits consideration for replication in other crop improvement programmes. Furthermore, RTIP's development of an informal network whereby it exploited the knowledge and resources of a host of institutions and resource persons throughout the country and beyond provides a model that could and should be replicated by other projects and institutions. RTIP collaboration with IITA and the exchange of experiences with IFAD projects on roots and tuber crops in the region are also excellent features that merit replication.

The performance of the central partners in the basic tasks of project design and implementation, including loan administration, technical backstopping, and implementation support has been without any major problems. The main weakness in IFAD performance was in the design of the project with its insufficient or incomplete approach to targeting and its choice of an unsuccessful strategy for covering processing and marketing questions. To its credit, it has added intellectual and financial resources to complement the supervision process by the cooperating institution. The physical presence in country of the World Bank as the cooperating institution has been a strength in terms of accessibility by the project and resultant efficiency of loan administration. However, this has also had drawbacks where the Bank has become over-involved in budget details of individual project activities and expenditures. Its took an interesting and innovative "team" approach to the implementation support dimension of supervision that brought benefits to RTIP by exposing it to a wide range of knowledgeable experts. A drawback of this approach that would require correction was that responsibilities and accountability were sometimes too diffuse.

The International Institute for Tropical Agriculture (IITA) proved to be a very valuable partner in providing technical backstopping to RTIP through an IFAD technical assistance grant. Their involvement was a successful feature of project design and of project implementation. The fact that IITA has also supported other IFAD efforts in root and tuber crops in the region was an added benefit that has facilitated even further sharing of knowledge and exchange of experiences. The performance of RTIP's limited number of NGO's and CBO's partners was good and RTIP benefited from their outreach capacity and knowledge of local conditions.

The overall assessment of the evaluation team is that, given its ambitious size and its complexity, RTIP has been well-managed and successfully implemented with good results in most areas. Its greatest strength was in creating a well-functioning nationwide system for multiplication and distribution of improved planting materials. In the process it strengthened numerous institutions, increased farmer access to information, improved production practices and contributed to better household food security. By contrast it fell short of its full potential in terms of poverty reduction impact, due in part to deficiencies in the RTIP's design and in part to RTIP's incomplete or inadequate implementation of its planned activities.

The most serious design deficiency was what proved to be unrealistic and unsuccessful provisions for covering post-production and marketing issues. This design deficiency was compounded by inadequate attention to these same issues in implementation. Of comparable gravity was the deficiency in RTIP's design with respect to targeting. RTIP lacked provisions that could and should have been made to take a "pro-poor" approach in each of its components. It also lacked measures to ensure a better gender balance in the accrual of benefits. As a result, during RTIP implementation, management and technicians put technical considerations before target group considerations, meaning that impact on farmer incomes was not taken adequately to heart. During implementation RTIP also over-emphasized cassava with relative neglect of yam and cocoyam that meant neglect of the zones and households that favour those crops.

The recommendations for the remainder of the RTIP implementation period and the preparation of future investments are:

• Limitation of further distribution of improved cassava planting materials to areas with proven high demand for cassava for processing, i.e. PSI areas,
• Completion of IPM infrastructure and conducting a study to assess financial and economic impact of RTIP IPM approaches
• Continuation of varietal release programme, with production of extension materials and the development of a web-site to store and share information generated by RTIP
• Expansion of Adaptive Research component to include other research needed to achieve RTIP objectives
• Concentration of support to groups on processing groups and producer groups working with the Ayensu Starch Company
• Recruitment of 1-2 officers for post-production and marketing component to review existing situation and help identify needs in an eventual Phase II investment
• Review of the current M&E system to pare down collection and processing of data on RTIP physical and financial progress

The Office of Evaluation supports a second phase investment to follow up and consolidate investments and activities that have been undertaken through RTIP to-date. In this context its recommendations are the following:

• Future investments by the Government of Ghana in support of the development of particular crops, as in RTIP, should be designed to support the entire vertically integrated commodity chain.
• Future IFAD investments in the development of root and tuber crops in Ghana should maintain improved food security and increased household incomes for the rural poor as their overall goals.
• Past assumptions about crop sector development and its impact on poor rural households should be carefully re-examined in the design and in the implementation of future investments.
• Large scale campaigns for the multiplication and dissemination of improved varieties should be used to reduce rural poverty only once socio-economic benefits to poor farmers who adopt improved materials have been fully assessed and only in cases where it is clear that the market can absorb a large supply response, as with the government PSI initiative.
• Agricultural research and farmer field schools, or other extension activities, should be demand-led. Farmer priorities must be routinely ascertained and given ample weight alongside technical and agricultural policy considerations.
• A study should be undertaken of the comparative costs and benefits of farmer field schools versus those of normal extension practices at the district level before further investments are made in this approach.
• Future investments for reducing poverty through investment in root and tuber crops should emphasise activities related to post-harvest, marketing and development of new market opportunities.
• Among post-harvest and marketing activities that should be considered for support are: (i) appraisal of technical and financial viability of existing processing equipment; (ii) training and advisory services on processing techniques; (iii) training on hygiene, health and environmental issues at processing sites; (iv) advisory services on packaging and labelling; (v) improved storage methods, (vi) regular dissemination by radio of price information; (ix) promotion of linkages between producers, processors, and traders on outputs and equipment and, (vii) elaboration of various financing models and arrangements with financial institutions to fund processing equipment and working capital requirements.

---

1/ This study was done by the Ghana Institute for Management and Public Administration, (GIMPA).

2/ See Appendix 1

3/ See Appendix V (page 59) for further information on economic and food security aspects of the RTIP.