The aim of the Interim Evaluation Mission (IEM) was to review the implementation of the Toledo Small Farmers Development Project (TSFDP) in relation to the stated hypotheses, strategies, targets and budget; to identify the crucial factors influencing implementation; and to evaluate the degree to which project objectives have been achieved. On the basis of implementation experience the IEM was to recommend whether a second phase to the project was warranted, and, if so, to propose future project concepts, strategies and broad design features.

Background

The Toledo district, which is the most southerly of Belize’s six main administrative units, is bordered on the west and south by Guatemala and to the east by the Caribbean Sea. Of its total area of 4 650 km2, about 100 000 ha are suitable for annual crops, fruit trees and pastures. The district is composed of flat coastal plains, gently rolling foothills and undulating lowlands, and the relatively high Maya Mountains reaching about 1 000 metres in elevation. Four main rivers, which are noted for rapidly rising water levels during the wetter season, make access to the southern half of the district by land almost impossible. Main access to Belize City and Belmopan, the capital, is by road (340 km), half of which is unsealed.

The uplands are classified as wet sub-tropical, with annual precipitation of 3 000 to 4 500 mm, while the lowlands, with a yearly rainfall of 2 000 to 3 500 mm, have a wet tropical climate. Severe storms, which sometimes significantly damage crops, occur in the wetter season (June to November). Soils are quite deep in the lowlands but generally shallow in the hill areas. In south-western Toledo, where farmers are concentrated, soils are quite fertile.

The total population in the Toledo district was about 14 100 in 1985 and 17 500 according to a 1991 census. Of the 2 710 households estimated in 1985, some 78% were rural, living in villages and settlements ranging from 12 to 430 persons. The largest group, the Ketchi Maya, increased from 32% to nearly 41%, while the second largest group, the Mopan Maya, decreased from 25% to 22% from 1985 to 1991. This change in proportions of the two groups may have been affected by refugees and economic immigrants predominantly of Ketchi Maya extraction settling in the project area. In 1991 the Mestizo, Garifuna, East Indian and Creole ethnic groups represented 11.9%, 10.0%, 7.9% and 5.7% of the population, respectively.

Project design

The Staff Appraisal Report (SAR) stated that in Toledo district, a priority area in the Government of Belize's (GOB) overall plans, the main constraints to agricultural production were poor marketing, lack of transport infrastructure, inadequate services and the unavailability of inputs. With specific reference to the Maya Indians, the SAR noted that shifting subsistence slash-and-burn cropping was practised and that with increasing land pressure, it was no longer sustainable. The project was designed to replace the shifting cultivation, which prevailed among small farmers, by demonstrating and assisting with the implementation of improved, stabilised farming systems. Technical packages of proven practices, some of which were already in use within the district, were said to be adequate and available.

Basically the project was to focus on providing the small farmers with access to essential services and improved production technologies. This was to be achieved by improving government services in the areas of farm credit, extension, input supplies and marketing facilities. To improve services the project would support increases in staff numbers, staff training, better extension-research coordination, the establishment of retailing and marketing depots and the provision of credit.

The basic objectives of the project were summarised in the SAR as follows:

to improve the income and consequently the standard of living of a group of small, mostly subsistence farmers now living in isolation, by bringing them into modern agriculture;

• to increase their agricultural production and productivity, as well as their involvement in the economic and social activities of the country;
• to strengthen the institutions which provide support to the farmers, including agricultural extension and credit; and
• to augment the physical infrastructure for handling, processing and marketing the main crops.

Of the 2 110 rural households in 1985, an estimated 1 620 or about 78% were considered to be small farmers, a number which was expected to increase to about 1 700 by the time the project commenced, according to the SAR. The farm income of the target group was estimated at USD 1 000 per household per year or USD 175 per capita. Child mortality was 66.5 per 1 000, which was much higher than the national average of 40, while infant malnutrition in Toledo district at 49% was the highest in Belize.

The SAR was originally conducted in 1985. Because of the need to coordinate with the USAID financed project, the Toledo Agricultural Marketing Project (TAMP), the loan agreement was signed in 1987 and the project did not become effective until August 1988. Its envisaged five-year duration was extended twice, initially until October 1993 and later until April 1995 (activities to be closed by October 1994). The project's total budget was some USD 2.9 million, plus USD 1.0 million through TAMP. IFAD's original contribution was approximately USD 2.2 million.

Main project outputs

The project’s performance has been quite erratic. As noted in the Mid-Term Evaluation (MTE) in January 1992, the project had a very slow start-up, with little being achieved before 1991 and a successive build-up in loan allocation and disbursements once several misconceptions in project design had been overcome. It also was characterised by variable performance in staffing, including a permanent lack of extension officers, insufficient experience of SMSs and excessively frequent turnover of project technical staff. The project's major outputs are summarised in the following paragraphs.

Expenditure. According to IFAD records, 75% of project funds had been disbursed by the end of June 1994, compared to 41% at the time of the MTE conducted in January 1992. The overall disbursement is relatively poor, considering that project life has been extended by two years. The low disbursement in the credit category (43%) which amounts to 31% of the project’s funding, together with an under-utilisation of funds for M&E and consultants services, has heavily affected overall disbursement.

Facilities. Although there were delays, the project infrastructure was built as stipulated in the SAR. It built two marketing depots, a feedmill and livestock holding facilities, took over facilities established by a previous project at Blue Creek as the project’s operational centre, and provided housing for project staff and extension officers.

Staffing. The project clearly suffered from a lack of direction until the appointment of a new PD in March 1991, after which project performance picked up. In terms of staff technical assistance (TA), the positions stipulated in the appraisal were filled, except for the Accountant (not filled until 1991) and the Marketing Specialist. The situation with counterpart staff for the project's Farm Systems Specialist and Livestock Specialist provided under TA, and Extension Officers (EOs) who were to be provided by the Ministry of Agriculture (MOA), was variable. After some delays counterpart staff were provided, but their involvement in project activities diminished after the departure of the TA specialists, and for a period in 1993 both counterpart positions were unfilled. The MOA at no time staffed the district adequately with EOs. Overall, staffing (both in number and quality) has been one of the least positive project features.

When it became apparent that the MOA was unable to provide adequate advisory services, the project, in response to a suggestion by OPS, instigated a Village Farmer Helper (VFH) scheme. Key farmers in remote areas, chosen by their peers and the project/MOA, received some training and are expected to provide advice to their fellow villagers in the future. This methodology, provided the VFHs receive adequate support, has several advantages, i.e.: it could help overcome the MOA staffing problems; through feedback from VFHs, it could also provide a better understanding of farmers' aspirations and limitations; thirdly, it could help overcome communication problems related to language barriers or to the disinclination by villagers to cooperate with government officials. Progress with the VFH scheme is encouraging, but better training as well as continuing material and technical support will be needed.

Agricultural Development. The project has made progress toward its objectives of increased production and encouragement of sedentary agriculture through the provision of advice, better genetic material, credit and the allocation of surveyed plots of land to farmers under individual title. Activities not entailing borrowing from TSFDP-provided funds are difficult to quantify, particularly as the project's M&E component was almost totally ineffective. Enterprises requiring credit (especially for medium- to long-term development loans) are easier to quantify.

Over 1 600 farmers attended training courses from 1991 to 1994, on topics as diverse as the safe use of chemicals, livestock husbandry and land preparation for rice. The project also procured and distributed improved strains of poultry, pigs and cattle, as well as some seed throughout the district. A kitchen vegetable garden scheme recently promoted by the project now involves women in 12 villages and has the potential to further involve women in development. The impact of training on production is uncertain, except for rice.

Data available to the IEM showed increased crop production only in the case of rice. As it evolved, the project concentrated on rice with 83% of loans (in number) and 67.3% of the loan value going to rice-based enterprises. By facilitating the surveying and allocation of over 230 blocks of land and providing more than 100 development loans for rice (together with about 130 crop loans), the project contributed to the expansion of mechanized rice in Toledo district, where estimated total rice production (paddy) increased from 3 137 tons in 1988 to 4 113 tons in 1993.

GOB price policy on rice also played a major role in such production increase. Until 1993/94 the Toledo rice producers were in a relatively favourable position, since rice production in Belize did not meet domestic demand and BMB was paying a subsidised price, on average, of over B$ 0.22/lb paddy which was well above the world price equivalent. However, if national production in 1994/95 rises to over 14 000 tons compared to 9 600 tons in 1993/94 (BMB estimates), it is conceivable that Toledo will have trouble selling rice. The GOB is faced with a serious dilemma on how to dispose of a potentially huge surplus of rice and how to deal with the increasing burden on its budget due to its price policy on rice.

The uncertain sustainability of the current level of price subsidies on rice has led the IEM to be very cautious about the present success of TSFDP in increasing production of mechanized rice by concentrating most of its resources on this investment. It would appear from the data presented to the IEM that few mechanised rice producers could withstand a substantial drop in price. With lower prices, rice cultivation can only remain profitable with high yields. Yet, weeds and a drop in the amount of fertiliser used are likely to become limiting factors, making it necessary to fallow or rotate land; in addition, the issue of drainage is still unsolved. If mechanized rice production becomes unprofitable, the repayment of a large share of DFC portfolio under TSFDP will be in jeopardy.

At this stage suitable alternative enterprises have not been confirmed as financially viable. The 37 citrus and 18 cattle loans are still in the development stage, therefore they have not contributed to production as yet. Their financial outcome with respect to viability and sustainability is still uncertain.

Conclusions

Achievement of project objectives

The primary objective of the project was to improve the income and consequently the standard of living of a group of small, mostly subsistence farmers by introducing new agricultural practices and inputs. Thus their agricultural production and productivity would be increased and they would be more involved in the economic and social activities of the country.

Assuming that the granting of a legal lease on land and approval for development loans are valid indicators of a shift from slash-and-burn farming to sedentary agriculture, by June 1994 the project had reached a maximum of 235 households. Since some of these leases have undoubtedly gone to farmers already engaged in sedentary farming (and others, including public servants and businessmen from Punta Gorda), the proportion of farmers within the target group that has been introduced to settled farming is small. Nevertheless, the project has kindled interest among other shifting agriculturalists to obtain their own individual leases, if not for increased productivity, at least to have their own land. Therefore, the project has introduced a relatively small number of farmers to sedentary farming and raised the possibility of more following the same path.

There is little information on whether the income and standard of living of project beneficiaries has actually been improved, partly due to a lack of monitoring by the project. Given the current level of subsidies on rice, mechanised rice for farmers achieving yields of 2 500 lb/acre or more are profitable, and some 230 loans (development and crop) have been granted for this activity. It is too early to assess the impact of cattle and citrus loans, but the prognosis (for the latter certainly) is not encouraging.

A second objective was to strengthen the institutions which provide support to the farmers, including agricultural extension and credit. In this respect the IEM's main conclusions are as follows:

• while the project enhanced the operational capacity of the district's MOA, its effects with regard to training and improvement in the mode of operation appear limited;
• the build-up of the DFC's institutional capacity has benefitted from the provision of an additional staff member and better operating means. Yet, in spite of the excellent performance of its staff in Punta Gorda, the DFC still suffers from a limited capability to operate effectively in servicing loans, providing technical advice and supervising loan repayments in the field; and
• the project enhanced the Ministry of Natural Resources' operational capability and its ability to relate to farmers.

The project's third objective was to develop the physical infrastructure for handling, processing and marketing the main crops. While providing most of the stipulated infrastructure, the project's achievements fell far short of its aspirations. Aside from the organic cacao development, the project has achieved virtually nothing in the way of identifying produce markets and marketing channels. The failure of the GOB to sanction the employment of a marketing specialist has caused serious setbacks in the project.

On the positive side, the project has helped coordinate marketing of rice through the Belize Marketing Board's milling facility and appears to have defined a niche to market "organic" cacao. It established two marketing/input supply depots and has trained personnel in the management of these, in addition to the one built by TAMP. The depots have essentially served as input suppliers and have not achieved any marketing function. Under the guidance of the project, management of the depots is improving, but their viability is tenuous, given their presently limited function.

Relevance of project design

The original design was based on a number of assumptions, some of which proved to be valid, e.g. the following:

• that Toledo district was a marginal area where most inhabitants (mainly Maya Indians) were eligible for IFAD assistance;
• that the traditional slash-and-burn cultivation system was not sustainable in the long run and had to be modified; and
• that the demand for credit was important enough to launch a credit program as an essential part of the TSFDP.

Other SAR assumptions were not validated during the course of implementation. These include the following:

• that the technology for modifying the existing slash-and-burn system was available and ready for dissemination to the project's target group. This proved incorrect and, in what appeared to be a conscious policy decision, the project switched much of its efforts to mechanised rice farming (where it was possible) as a first step in sedentary farming;
• that the two cash crops promoted, cacao and annatto, were financially attractive. The opposite proved to be the case, due to changes in world prices and the absence of internal mechanisms of price stabilization;
• that DFC, as it was structured, was ready to be used as TSFDP's credit delivery system. While it was essentially correct to assume that credit was a major bottleneck to development, the design did not appear to be aware of the consequences on targeting objectives implied by the need for legal individual access to land as a collateral for borrowing from DFC for development loans. Given land tenure arrangements on (real or supposed) Maya reservation land, the SAR did not consider the target group's need of credit delivery, supervision and management mechanisms to be compatible with traditional socio-economic structures and communal land ownership, especially on the hills; and
• that TAMP and TSFDP were to complement each other in marketing development with some critical interaction between staff members. In reality, the envisaged cooperation never took place and the marketing components of both projects fell short of their goals.

The SAR assumed that the farmers belonging to the target group were aware of increased population pressure on land and therefore, were prepared to change their farming systems. While an objective assessment indicates that currently much of the target area cannot sustain a cycle of seven years' fallow followed by one year's cropping, farmers appear to be far from convinced that their traditional method of farming is being eroded because of land pressure. However, as a result of the project, some villages promoted land titling on what the villagers considered their communal land, and are beginning to realize that land shortages are becoming a reality.

With regard to coordination, the SAR assumed that through the mechanism of the National and District Coordinating Committees, without any single line of control, the district's MOA and the project could coordinate their activities and act essentially as one. This did not always prove to be the case. It was probably unrealistic of the design to expect a degree of cooperation and coordination without a single authority being ultimately responsible for implementation.

The assumption that DFC could effectively reach and service small farmers, whose cash requirements for inputs were small, was too optimistic. DFC would not lend without land security and, for reasons of costs, it was not interested in small loans to individuals. Other lending mechanisms were obviously needed if the bulk of the target group were to be reached.

Sustainability

The sustainability of the programmes and activities established and promoted under the project is likely to depend on the future performance of marketing, credit, advisory services and, with regard to subsidies (particularly on rice), of government policies. The identification and promotion of new and modified farming systems will be critical to attain sustainable agricultural production.

Recommendations

Because the IEM is convinced that population pressure eventually will impose substantial changes to traditional shifting agriculture, and that enterprises established under the project need further support, a second phase of the project is recommended but only with major shifts in emphasis and only under certain conditions. Because of the numerous problems identified in the course of implementation of the first phase, IFAD should negotiate procedures and means as to avoid the recurrence of the same problems in the future. Some guidelines are given by the IEM in this respect, as specified below.

General recommendations

Extreme care should be exerted when attempting to incorporate fundamental changes to existing farming systems. The assumption must not be made that changes appearing necessary and logical to planners coincide with the views and aspirations of the target group. Instead, a programme of field problem analysis, applied research and field demonstrations should be undertaken in consultation with the target group at all times. A time-frame of at least seven years should be allowed for such a programme.

Before allocating implementation responsibilities, the mandate of the organisation selected should be carefully scrutinised to see if it has the means to carry out the proposed tasks. For instance, DFC required land collateral and was ill-equipped and unenthusiastic about making small loans to small farmers since the beginning of TSFDP.

The project should not give excessive significance to cooperation and coordination between implementing agencies without the presence of a single line of authority. The MOA was designated as the implementing agency, with responsibility for the overall coordination of project activities. Yet, it was unable to assure an adequate level of coordination between the TSFDP and the other implementing agencies. As regards extension activities, the IEM recommends that such inability be avoided by unifying the two functions of project director and district's MOA into a single line of authority.

Follow-up project

The second phase should focus on the farmers in the foothills and hills, but should also develop assistance to farms supported during the first phase, as a period of consolidation will be necessary. Should IFAD management approve a second phase, it should be preceded by a number of essential studies and clarifications by GOB on several issues (see below, section 2.a) and it would include some key issues, as explained below (section 2.b).

(a) Pre-project activities

The following studies are recommended prior to final project preparation (these could be conducted as part of a further extension to phase one, should it be granted) and approval:

• Marketing. This critical activity represents the marketing TA that was not implemented in phase one. The study would aim at the identification and appraisal of marketing opportunities and mechanisms; it would form the linchpin for future activities. Given that the Toledo district is disadvantaged, particularly with regard to communication, a specialist, with the help of the MOA and commercial organisations, should clearly identify any advantages that Toledo has in enterprise development. Conversely, it should rigorously exclude enterprises where the district has a comparative disadvantage. The assessments, while focusing on the needs of the Toledo district, should not be carried out in isolation but should take into account the overall policies that the GOB is reportedly in the process of developing;
• Rice Policy. The existing rice policy should be reviewed, with particular reference to a number of elements including production patterns (milpa and mechanized rice), future price policy, national self-sufficiency, quality control, technologies used, and impact of actual/proposed policies on the different regions and ethnic groups of Belize. Inter alia, an action plan for rice production by small farmers in Toledoshould also be defined by GOB;
• Credit. In order to assess the effectiveness and health of its portfolio in Toledo, DFC should carry out a survey on the impact of its credit activities (including technologies used and recovery policy) on the productivity and income of beneficiaries;
• DFC. An assessment is needed of whether DFC, given its mandate, can realistically lend to small farmers, particularly individuals without legal access to land. Should DFC prove unequal to the task of lending to small farmers, other mechanisms (i.e. NGOs) should be sought;
• Teaching/Research Facility. While the project headquarters will be the project-sponsored office block near Punta Gorda, a detailed study into expanding the current project headquarters at Blue Creek into a training and research facility is recommended. The purpose of the centre should include: (a) agricultural training for school drop-outs, farmers and their children, MOA staff and, in particular, Village Farmer Helpers (VFHs), and (b) a small research facility to serve as a base for on-farm investigations to be promoted during phase two;
• Drainage. As poor drainage is likely to limit other agricultural activities on a significant proportion of the land currently used for mechanised rice, a clear picture of the actual situation is needed. A study on the implications of poor drainage, methods and costs of improving drainage is recommended;
• Roads/Civil Engineering. As the IEM is recommending improved access to remote areas through a modest road programme, a plan with cost estimates and priorities of roads suggested should be prepared; and
• Monitoring and Evaluation. An M&E system appropriate to the needs of a phase two project, that can be easily handled by project personnel, should be developed. A study for this purpose, which will include proposals on staff training, is recommended. The possibility of contracting a non-government agency for M&E purposes should be considered given the past poor performance of the institution in charge of the implementation of this component (MED).

(b) Phase two: project proposals

The components of a phase two and the manner in which they are implemented will depend to a large extent on the pre-project investigations now being proposed by the IEM, as well as on the preliminary negotiations that IFAD should have with the GOB before the appraisal of a potential second phase (e.g. on inter-agency coordination, quality and quantity of staffing, political commitment of GOB to project objectives, credit delivery, extension, etc.). They will, however, contain some key elements which are summarised in the following paragraphs.

As an overall strategy, the second-phase project would use the village’s social structure to promote development activities and assure that the views of the village are fully recognised and taken into account in project design, rather than concentrate on individual households.

The project would concentrate on promoting development activities in the foothills and hills, where lives most of the target group that was not reached by phase-one TSFDP, while consolidating activities in foothill and lowland areas that already benefited from the project services.

To promote development in the foothills and hills, it is recommended that the following measures be adopted:

• use adequate technical assistance to identify alternative farming systems suited to hillsides and test them on site as necessary. It is essential that a farm system and a village approach, and not an individual crop and household approach, be adopted;
• to ensure that identified markets, credit delivery channels and mechanisms, and mainly technologies can be adequately tested and promoted, a project life of seven years, with a thorough interim evaluation after two years, is recommended;
• develop a technology transfer system including the village community as a whole through the VFH system as well as individual farmers as necessary;
• considerably extend the village household garden programme, both because of its potential to improve household nutrition and to provide a stepping-stone for future activities;
• in order to facilitate socio-economic development in remote villages, initiate a road improvement programme; and
• retain the TSFDP facilities at Blue Creek as a training and modest research centre.

If positively assessed by the second phase MTE, the VFH system should be extended to other districts after the VFHs are properly trained at the TSFDP Blue Creek Centre.

The project would promote credit mechanisms which are more "user-friendly" than those currently used by DFC, but only for activities proven to be viable. Supervision of all loans provided under the project would be actively promoted. The project would elicit cooperation between DFC (or another agency if necessary) and the village community in promoting and supervising activities and encouraging loan repayments.

The farming activities established by TSFDP would be supported and/or modified as necessary. The longer term future of both rice and citrus appearing precarious, the project would endeavour to increase the profitability of those enterprises by promoting better crop husbandry, while at the same time instigating alternative complementary activities. In this manner rice, as a mono-culture, would be de-emphasised.

Should the recommended studies confirm economic feasibility, the project would assist in the establishment of a small citrus processing facility and promote a drainage network wherever appropriate.

Subject to economic and environmental sustainability, livestock production should be promoted, partly through the financing of an abattoir.

Using the strategies developed in the recommended study looking at developing the facilities at Blue Creek as a training/project research centre, the project would concentrate on developing the VFH extension concept more fully. To effectively do this, training would be critical. Technical assistance would be used as necessary to train Subject Matter Specialists (SMSs) and Extension Officers (EOs) both in technical matters and extension methodology. The VFHs would be trained by the EOs and SMSs with some direct input from TA.

As far as project organization and management is concerned, the following approach is recommended:

• the MOA would have overall responsibility for coordinating project implementation;
• at the district level, the duties of Project Director (PD) and District Agricultural Officer (DAO) would be carried out by the same person. This person would delegate much of his routine duties to a deputy DAO. In this manner the project would have a single line of control and be more likely to stay "on course". It is critical that the PD/DAO be a competent and dynamic senior person with similar capabilities to the TSFDP's current PD;
• the coordinating committees would be convened as necessary, generally at the request of the PD/DAO, and meet with supervising missions;
• a substantial amount of TA would be used for training, adaptive research and marketing. The TA personnel's terms of reference should be clear. Such personnel should not be overburdened with a multitude of tasks. They would report directly, and formally, to the PD/DAO; and
• GOB would prove its capacity to effectively monitor project activities along lines agreed with IFAD.

These proposals are essentially in line with the MOA's policy of introducing effective agricultural training programmes, developing an aggressive market information system and promoting agro-processing in the project area.