Completion evaluation

The completion evaluation of the Agricultural Credit Project for the Loan 163-ES, El Salvador, for which the IDB has been the Cooperating Institution, was carried out in November/December 1990 with the intention to draw lessons and provide specific recommendations for the Paracentral Regional Agricultural Development Project (Loan 263-ES), approved by the Fund in October 1990. The Agricultural Credit Project was approved in December 1984 and became effective in September 1985; it was closed in June 1990. The project made possible some increases in maize yields and the majority of beneficiaries reported an improvement in their general wellbeing. But the project results were negatively influenced by the price policy (which kept basic grain prices low), inadequate project supervision, deficient management at the government executing agency and the lack of an appropriate information system.

Background

During the nineteen-eighties, the economic and political situation of El Salvador worsened considerably, as a result of a combination of increasing internal problems, the acute crises of the Central American Common Market and the fall in world market prices for coffee and cotton, the country's main exports. Political violence resulted in a massive forced migration, about one million rural dwellers having to resettle in other parts of the country, and a further million to million and a half, migrating to other nations, chiefly to the United States and to neighbouring Central American countries.

Between 1980 and 1983, the gross national product fell in real terms and thereafter increased slowly, only reaching the 1980 level ten years later. Because political violence is concentrated in rural areas, the agricultural sector suffered most. The fall in export prices had a large negative effect, with the volume of export agriculture decreasing by 50% during the decade, while production for internal consumption did not increase, making it necessary to rely on increasing imports of food.

The agricultural sector also faced very important structural changes during the 1980's, as a result of new socio-economic policies. Among the most important of these was the agrarian reform, which even if it was not completed as originally planned, still resulted in a redistribution of 25% of all agricultural land. In the large expropriated haciendas, large production cooperatives were set up by the Government. In those areas where very small sharecroppers and renters farmed, the lands were sold to them at very low prices. Other important changes in the sector, consisted of the reform of the private banks and the State took over the marketing of export crops.

It was in this setting that the Agricultural Credit Project, partially financed through IFAD's loan 163-ES, began operating in 1985. This project was part of the III Stage of the Global Agricultural Credit Programme (PGCA), originally initiated in 1976 and which was jointly financed with the Government of El Salvador (GOES) and the Inter-American Development Bank (IDB).

Project objectives, basic assumptions and targets

The PGCA's basic objective was that of using the credit provided by the Agricultural Development Bank (BFA), to reactivate the agricultural sector, through loans to micro-, small and medium size farmers. The microproducers who obtained ownership through the agrarian reform's Decree-Law 207, were identified in the President's Report as IFAD's target group. It was proposed that the PGCA would be a supervised credit programme, with technical assistance to farmers coming from CENTA, FINATA and CDG. Other services, such as marketing, would be provided through existing private and public channels.

Food and export crops were to be incentivated through credit, leading not only to an increase in the volume of output and a 25% to 518 increase in income to participating farmers. Additional output was to result solely from increases of 30% to 180% in crop yields. This increase in productivity per hectare was considered feasible with the available technology and agricultural extension capacity.

The overall target of the PGCA was to provide credit to about 28 700 farmers over four years. In the case of IFAD's part of the programme, the target was to reach some 20 000 microproducers, which would have implied a total of some 61 655 subloans, including repeat financing of beneficiaries over the four years. Out of the total project cost of US$ 54 million, about US$ 7.6 million were assigned for subloans to microproducers. About 66% of these funds were to be provided by IFAD and the balance would come from IDB and BFA funds. It must be pointed out that both in the project appraisal document and the Loan Agreements, there is a certain ambiguity concerning the identification of funds by source for the different types of farmers. However, in IFAD's President's Report to the Executive Board, it is clearly stated that IFAD's target group would be 20 000 microproducers.

The project covered the whole of El Salvador (21 000 Km2) and the implementing institution was the BFA. The appraisal considered that BFA had the capability to handle the proposed 28 700 farmers, as it was at that time already providing services to 90 000 credit users. IDB participated as co-financier and cooperating institution for the supervision of IFAD's loan. Among the BFA's obligations in IFAD's Loan Agreement was that of setting up a monitoring and evaluation system to assist in project implementation.

Main results of the project

Overall, the project did not achieve its credit targets. Some 16 600 subloans were made with IFAD funds in six years (rather than the four originally planned). This represented about 30% of the original goal of 61 655 including repeat loans.

Concerning the use to which the subloans were allocated, in approximately 90% of the cases these went to cover production costs of grains as proposed at the design stage, the remaining 10% being allocated to farm investments. On the other hand, 76% of the funds were provided to microproducers, given that 19% of subloans went to other small farmers and an additional 5% to medium farmers.

Funds disbursed out of IFAD's loan were concentrated in just two years (1985 and 1987), with a very reduced use of IFAD resources in other years (1986, 1988 to 1990). This did not follow the original expectations at appraisal, that funds would be disbursed in an increasing progression through the third year, with a slight downturn in year four. Resources from subloans paid back were to be re-cycled for new lending through the project's revolving fund. Again, according to BFA reports, the use of the revolving fund resources was concentrated in just one year (1987), and large balances remained in the fund in 1986, 1988, 1989 and 1990, that were not used for providing new credit to the target group. This use of IFAD loan resources was due to BFAs's policy of making decisions at headquarters modifying the allocation of subloans made at agency level in order to use first available funds from the various sources (USAID, IDB, IFAD), rather than request disbursements of new funds from any source such as IFAD. Thus, there was no planned use of these funds, that would have ensured that the resources from each specific source would be available to finance a clearly identified group of clients during the project's implementation period.

The revolving fund resources were used for new loans in 1985 and 1987. Since 1988 these funds were mostly used to make provisions for bad debts and to cover foreign exchange losses.

Data corresponding to changes in crop productivity and incomes are very scarce. Nevertheless, some increases in yields can be inferred if data on project beneficiaries is compared with that of other micro- and small farmers. In the case of maize, which is by far the principal crop, credit users obtained yields of between 2.2 and 2.7 tons per hectare, which were higher than the 1.9 to 2.0 tons per hectare reported for other farmers (far below the irrealistic 3.5 tons per hectare that was set as the goal at appraisal). Regarding changes in income, there is no comparative data available concerning the beneficiaries situation before and after the project. Nevertheless, a survey carried out among IFAD credit users indicated that 58% of these reported an improvement in their general wellbeing, while 39% reported no change and only 3% indicated their situation had worsened.

The reduced number of beneficiaries reached by the project was directly related to BFA's administrative problems, which by not keeping separate accounts by source, allowed funds recovered from subloans that were destined for the revolving fund, to be used for other unspecified purposes by BFA. Changes of a positive nature in BFA management, which took place in mid-1989, came too late to benefit the project.

The results obtained by credit users in regards to increasing output and yields were negatively affected by the low prices for their products, lack of adequate farm technologies and insufficient support from the sectoral institutions that were originally assigned the role of providing technical assistance to farmers. On the basis of the available information, it can be stated that few farmers modified their production systems in order to increase their yields. On the other hand, it was observed that some of the credit beneficiaries employed the funds obtained to finance other economic activities (mostly of a non-agricultural nature) and other family expenditures (principally housing) that satisfied their more immediate needs.

Conclusions

Concerning project design: In order for its projects to have a greater positive effect on the target group, IFAD should participate in project design from its very beginning. Since the project being evaluated was part of the third-phase of an on-going global credit programme, project design was based on the experience with the previous phases which were directed to farmers with more resources. To incorporate micro- farmers in the new project, no substantial effort was made to establish what their needs were and how credit could be of use to them. The identification of the target group was not adequate, and this was a weakness of the design.

The project's proposal to channel most credit to the production of basic grains, while reflecting the type of crops which these farmers cultivate, meant that credit was being tied to the financing of low profitability crops. In addition, the assumptions relating to possible yield increases that were to be generated by the project were too optimistic. Furthermore, the initial decision to cover the whole country with the project generated serious implementation problems. The new 267-ES loan is restricted to a limited area.

The political and policy context: Political violence was a factor, but not the only one, affecting the project's normal implementation. The principal macroeconomic policies did not have a major positive effect on project development, as they focused mostly on export crops, while the project's emphasis was on crops for internal consumption. One policy that did have a significant negative effect was that of keeping basic grain prices low, as farmer beneficiaries produced mainly maize, beans and sorghum. Other negative effects were due to the frequent restructuring of the agencies dealing with agricultural research and extension, as the administrative changes hindered rather than improved services to farmers.

On project administration: Even though BFA had previous experience in managing credit projects with external funds, management of the IFAD funded project was not adequate. The special characteristics of IFAD's target group were not reflected by BFA in the credit regulations, which did not include the eligibility criteria that IFAD had proposed. Rather the credit regulations used BFA's own standard definitions for micro farmers and applied the same eligibility criteria it had always used. During implementation, BFA's procedure for assigning multiple subloans for a single farmer's credit operation, resulted in making it impossible to identify each client with any particular financing source. Thus IFAD's "beneficiaries" cannot be clearly established, as any farmer who received IFAD funds could also be receiving funds from IDB, or USAID, or any other source. This also led to project funds not being disbursed as planned at appraisal, but instead disbursements responded to the availability of funds in BFA from all sources. When sufficient funds were available from other sources, no request for IFAD disbursements were made; conversely, if IFAD funds were available then all loans from a particular agency that fitted the eligibility criteria were assigned to the IFAD project and a reimbursement request made. This, of course, meant that no planning of credit was possible and a continuous support for a particular farmer or farmers throughout the project period was again not possible. An additional serious problem derived from this situation, was that since IFAD funds were not handled separately from other resources, this led to reflows from the original subloans not being used for the granting of new credits.

The lack of a properly functioning information system during most of the implementation period, made project management more difficult and prevented the BFA from undertaking any systematic planning of credit use. The problems with the information system resulted also in monitoring and evaluation functions being superficially implemented, but without any real effect on project decision- making. However, it must be recognized that in the last three years significant advances were made concerning the information system, which with proper adjustments can become an important support for new projects.

On project supervision: The IFAD project was overshadowed by the much larger IDB contribution, which amounted to 75% of the total. The rapidly deteriorating financial situation of BFA seems to have involved the full attention of the Cooperating Institution on this problem, thus limiting IDB's interest in supervising what was happening with IFAD's funds. The reporting on project status by IDB to the Fund was insufficient. An additional problem was that several recommendations from IDB to improve project management were not generally followed by BFA. The Fund carried out some supervision missions but, mainly because of the lack of basic data concerning key implementation issues, these missions did not point out the need for immediate action (such as the threat to suspend loan disbursements, and, eventually their actual suspension or even the cancellation of the loan).

Recommendations for the paracentral region project

In October 1990, the Fund approved the Paracentral Region Agricultural Development Project (Loan 267-ES). In the design of this project several of the lessons learned from the Project 163-ES were incorporated. But, for the proper implementation of this new project some of BFA's present procedures need to be modified, in relation to operations policy, credit regulations and the type and frequency of reports generated by BFA's information system.

The BFA's administration has undergone substantial changes since mid-1989, which should be a positive factor for the new project's future development. However, a number of additional modifications in relation to operations policy are proposed which include that the three credit agencies participating (San Vicente, Sesuntepeque and Ilobasco), be considered as being part of a pilot project, in order to try, on an experimental basis, the following:

a) A savings mobilization scheme in their respective areas of operation, and that those savings be re-cycled to its clients, thus establishing a direct link between depositors and loan recipients. The additional credit resources needed to cover credit demand would be supplied by BFA's central office as occurs at present.

b) A major decentralization of the bank's operations, allowing each agency to establish its own balance sheets and accounts. This would allow each Head of Agency to manage his office's operations and to know the real cost of these.

c) To utilize the individual farmer as the basis for loan operations at the agency level, rather than considering each subloan with a different line of credit as a separate operation, like it is done at present.

Regarding the credit regulations, it is recommended that on the basis of experience with Loan 163-ES, IFAD should consider using as the main beneficiary eligibility criteria in El Salvador, the net worth of the loan recipient instead of his/her net income. The use of net worth would ease the BFA's work, since it already employs this criteria in its loan application reviews, and it also provides more trustworthy data on the economic status of the proposed beneficiary.

Finally, in relation to the information system, there is a need to introduce important modifications concerning the types of reports generated by the management information system (SIG). Only a few additional variables need be included in the existing information system. The main changes to be introduced into the SIG are:

a) The presentation of reports on credit operations, based on a standard report format, containing all the relevant data for project and agency management. A set of the proposed reports is included in Annex 5 of the report.

b) The elaboration of reports at the level of the three participating credit agencies, which could later be aggregated to produce a unified report for the project implementing unit, for BFA's headquarters and for the Cooperating Institution and IFAD. This could strengthen BFA's stated proposal to decentralize its activities.

c) Reporting on a regular basis would considerably assist in project monitoring, with some reports being produced weekly and others monthly, as needed. To fulfil report requirements of the Cooperating Institution, summary reports could be produced on a quarterly basis.

On the basis of the experience with the evaluation of the 163-ES project, it is recommended that the responsibility for the design and implementation of an evaluation methodology be assigned to PERA, an institution specialized in this type of activities. This unit is part of the Agricultural Sector Planning Office, which during 1991 will receive assistance in this area from the joint UNDP/IBRD/IFAD Regional Unit for Technical Assistance (RUTA) project.