Tunisia World Bank by alicejenny


									                                        Document of
                                       The World Bank

                                                        Report No.: 26260


                               REPUBLIC OF TUNISIA

                              (LOAN 3217)

                              (LOAN 3661)

                          (LOAN 3691)

                                (LOAN 3892)

                                       JUNE 26, 2003

Sector and Thematic Evaluation Group
Operations Evaluation Department
Currency Equivalents (annual averages)

Currency Unit: Tunisian Dinar (TND)
1990             US$1.00          TND 0.87
1991             US$1.00          TND 0.92
1992             US$1.00          TND 0.88
1993             US$1.00          TND 1.00
1994             US$1.00          TND 1.01
1995             US$1.00          TND 0.94
1996             US$1.00          TND 0.97
1997             US$1.00          TND 1.11
1998             US$1.00          TND 1.14
1999             US$1.00          TND 1.19
2000             US$1.00          TND 1.37

Abbreviations and Acronyms

AVFA            Agence de la Vulgarisation et de la Formation Agricole
                (Agricultural Extension and Training Agency)
BNA             Banque Nationale Agricole
                (National Agricultural Bank)
CRA             Centre de Rayonnement Agricole
                (Local Extension Unit)
CRDA            Commissariat Regional de Developpement Agricole
                (Regional Agricultural Development Commission)
CTV             Cellule Territoriale de Vulgarisation
                (Sub-Regional Extension Unit)
GDA             Groupement de developpement agricole
                (Agricultural Development Association)
GDP             gross domestic product
GTZ             Deutsche Gesellschaft fur Technische Zusammenarbeit
                (German Technical Assistance)
ICR             Implementation Completion Report
IRESA           Institut de Recherche et de l’Enseignement Superieur Agricole
                (Institute of Agricultural Research and Higher Education)
MNA             Middle East and North Africa Region, World Bank
MNSRE           Rural Development, Water and Environment Group of MNA
NGO             nongovernmental organization
ODESYPANO       Office du Developpement Sylvo-Pastoral du Nord-Ouest
                (North-West Forestry and Pastoral Development Agency)
OED             Operations Evaluation Department
PMU             Project Management Unit
PPAR            Project Performance Assessment Report
T&V             training and visit extension methodology
Fiscal Year

Government          January 1 – December 31

Director-General, Operations Evaluation               :   Mr. Gregory K. Ingram
Director (Acting), Operations Evaluation Department   :   Mr. Nils Fostvedt
Manager, Sector and Thematic Evaluation               :   Mr. Alain Barbu
Task Manager                                          :   Mr. John Heath

  OED Mission: Enhancing development effectiveness through excellence and independence in evaluation.

About this Report
     The Operations Evaluation Department assesses the programs and activities of the World Bank for two
purposes: first, to ensure the integrity of the Bank’s self-evaluation process and to verify that the Bank’s work is
producing the expected results, and second, to help develop improved directions, policies, and procedures through
the dissemination of lessons drawn from experience. As part of this work, OED annually assesses about 25 percent of
the Bank’s lending operations. In selecting operations for assessment, preference is given to those that are
innovative, large, or complex; those that are relevant to upcoming studies or country evaluations; those for which
Executive Directors or Bank management have requested assessments; and those that are likely to generate
important lessons. The projects, topics, and analytical approaches selected for assessment support larger evaluation
     A Project Performance Assessment Report (PPAR) is based on a review of the Implementation Completion
Report (a self-evaluation by the responsible Bank department) and fieldwork conducted by OED. To prepare
PPARs, OED staff examine project files and other documents, interview operational staff, and in most cases visit
the borrowing country for onsite discussions with project staff and beneficiaries. The PPAR thereby seeks to
validate and augment the information provided in the ICR, as well as examine issues of special interest to broader
OED studies.
     Each PPAR is subject to a peer review process and OED management approval. Once cleared internally, the
PPAR is reviewed by the responsible Bank department and amended as necessary. The completed PPAR is then
sent to the borrower for review; the borrowers’ comments are attached to the document that is sent to the Bank’s
Board of Executive Directors. After an assessment report has been sent to the Board, it is disclosed to the public.

About the OED Rating System
      The time-tested evaluation methods used by OED are suited to the broad range of the World Bank’s work.
The methods offer both rigor and a necessary level of flexibility to adapt to lending instrument, project design, or
sectoral approach. OED evaluators all apply the same basic method to arrive at their project ratings. Following is
the definition and rating scale used for each evaluation criterion (more information is available on the OED website:
      Relevance of Objectives: The extent to which the project’s objectives are consistent with the country’s
current development priorities and with current Bank country and sectoral assistance strategies and corporate
goals (expressed in Poverty Reduction Strategy Papers, Country Assistance Strategies, Sector Strategy Papers,
Operational Policies). Possible ratings: High, Substantial, Modest, Negligible.
      Efficacy: The extent to which the project’s objectives were achieved, or expected to be achieved, taking into
account their relative importance. Possible ratings: High, Substantial, Modest, Negligible.
      Efficiency: The extent to which the project achieved, or is expected to achieve, a return higher than the
opportunity cost of capital and benefits at least cost compared to alternatives. Possible ratings: High, Substantial,
Modest, Negligible. This rating is not generally applied to adjustment operations.
      Sustainability: The resilience to risk of net benefits flows over time. Possible ratings: Highly Likely, Likely,
Unlikely, Highly Unlikely, Not Evaluable.
      Institutional Development Impact: The extent to which a project improves the ability of a country or region
to make more efficient, equitable and sustainable use of its human, financial, and natural resources through: (a)
better definition, stability, transparency, enforceability, and predictability of institutional arrangements and/or (b)
better alignment of the mission and capacity of an organization with its mandate, which derives from these
institutional arrangements. Institutional Development Impact includes both intended and unintended effects of a
project. Possible ratings: High, Substantial, Modest, Negligible.
      Outcome: The extent to which the project’s major relevant objectives were achieved, or are expected to be
achieved, efficiently. Possible ratings: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately
Unsatisfactory, Unsatisfactory, Highly Unsatisfactory.
      Bank Performance: The extent to which services provided by the Bank ensured quality at entry and
supported implementation through appropriate supervision (including ensuring adequate transition arrangements
for regular operation of the project). Possible ratings: Highly Satisfactory, Satisfactory, Unsatisfactory, Highly
      Borrower Performance: The extent to which the borrower assumed ownership and responsibility to ensure
quality of preparation and implementation, and complied with covenants and agreements, toward the achievement
of development objectives and sustainability. Possible ratings: Highly Satisfactory, Satisfactory, Unsatisfactory,
Highly Unsatisfactory.

Principal Ratings................................................................................................................v

Key Staff Responsible ........................................................................................................v

Preface.............................................................................................................................. vii

1. Background ................................................................................................................ ix

2. Agricultural Services ...................................................................................................3

3. Rural Water Management ..........................................................................................6

4. Area Development........................................................................................................8

5. Rural Finance .............................................................................................................11

6. Ratings ........................................................................................................................14

           Outcome .................................................................................................................14
           Sustainability and Institutional Development........................................................16
           Bank and Borrower Performance..........................................................................17
7. Findings and Lessons.................................................................................................19

           Findings for Tunisia...............................................................................................19
           Lessons of General Relevance for Bank Operations .............................................21
Annex A: Agriculture and Rural Development Overview of Tunisia.........................23

Annex B: Results Matrices..............................................................................................25

Annex C: Supplemental Tables ......................................................................................31

Annex D: Basic Data Sheets ............................................................................................36

Box 1. Village Development at Traia, Bizerte ...............................................................10

Box 2. Response from the Region (MNSRE) to OED’s Comment in Paragraph 7.4 20

Table 1. Evaluation Design................................................................................................2

Table 2. How the Outcome Rating is Derived ...............................................................15

Principal Ratings
                          Outcome             Institutional         Sustainability     Bank                Borrower
                                              Development                              Performance         Performance
Agricultural Research and Extension Project
ICR*                      Satisfactory        Modest                Likely             Satisfactory        Unsatisfactory
ES*                                           Modest                Unlikely           Satisfactory        Unsatisfactory
PPAR                      Unsatisfactory      Modest                Unlikely           Unsatisfactory      Unsatisfactory
Agricultural Sector Investment Loan
ICR                       Satisfactory        Substantial           Highly Likely      Satisfactory        Satisfactory
ES                        Satisfactory        Substantial           Highly Likely      Satisfactory        Satisfactory
PPAR                                          Substantial           Likely             Satisfactory        Satisfactory
Northwest Mountainous Areas Development Project
ICR                       Satisfactory         Substantial          Likely             Satisfactory        Satisfactory
ES                                            Substantial           Likely             Satisfactory        Satisfactory
PPAR                      Satisfactory        Substantial           Likely             Satisfactory        Satisfactory
National Rural Finance Project
ICR                       Unsatisfactory      Substantial           Unlikely           Unsatisfactory      Unsatisfactory
ES                        Unsatisfactory      Modest                Unlikely           Unsatisfactory      Unsatisfactory
                          Highly                                                       Highly              Highly
PPAR                                          Neglibible            Unlikely
                          Unsatisfactory                                               Unsatisfactory      Unsatisfactory

* The Implementation Completion Report (ICR) is a self-evaluation by the responsible operational division of the Bank.
The Evaluation Summary (ES) is an intermediate OED product that seeks to independently verify the findings of the ICR.

Key Staff Responsible
                               Team Leader                       Sector Manager                Country Director
Agricultural Research and Extension Project
Appraisal                      J.P. van der Veen                 Rory O’Sullivan               Kermal Dervis
Completion                     Charles Ameur                     Mark D. Wilson                Salah Darghouth (Acting)
Agricultural Sector Investment Project
Appraisal                      Stephen D. Mink                   Odin K. Knudsen               Harinder Kohli
Completion                     Shobha Shetty                     Doris Koehn                   Christian Delvoie
Nothwest Mountainous Areas Development Project
Appraisal                      Mathewos Woldu                    Odin K. Knudsen               Harrinder Kohli
Completion                     Idah Pswarayi-Riddihough          Petros Aklilu                 Christian Delvoie
National Rural Finance Project
Appraisal                      Bernard Dussert                   Odin K. Knudsen               Daniel Ritchie
Completion                     Stephanie Gober                   Dean N. Jordan (Acting)       Christian Delvoie

This is a Project Performance Assessment Report (PPAR) of four projects in the Republic
of Tunisia:

   •   The Agricultural Research and Extension Project, for which Loan No. 3217-TU in
       the amount of US$17 million equivalent was approved on June 5, 1990. The loan
       was closed on June 30, 1997, one year behind schedule. The final disbursement
       took place on October 31, 1997, at which time a balance of US$1.7 million was

   •   The Agricultural Sector Investment Project, for which Loan No. 3661-TU in the
       amount of US$120 million equivalent was approved on November 18, 1993. The
       loan was closed on December 31, 2000, eighteen months behind schedule. An
       undisbursed balance of US$1.3 million was rolled into the follow-on sector
       investment project.

   •   The Northwest Mountainous Areas Development Project, for which Loan No.
       3691-TU in the amount of US$27.5 million equivalent was approved on
       December 23, 1993. The loan was closed on June 30, 2001, one year behind
       schedule, and a balance of US$1.3 million was canceled.

   •   The National Rural Finance Project, for which Loan No. 3892-TU in the amount
       of US$55.0 million equivalent was approved on May 23, 1995. The loan was
       closed on June 30, 2001, twenty-one months behind schedule, and a balance of
       US$7.6 million was canceled.

The PPAR presents the findings of a mission by the Operations Evaluation Department
that visited Tunisia in February-March 2003. The mission was conducted by Mr. John R.
Heath. The findings draw on interviews with beneficiaries, project staff, officials of the
Government of Tunisia and Bank staff. The collaboration of these persons is gratefully
acknowledged. In addition, the PPAR draws on appraisal documents, implementation
completion reports, and other background data.

This cluster of projects was selected to provide background for the agriculture and rural
development section of OED’s FY04 Country Assistance Evaluation for Tunisia. The
report incorporates data that was not available when the completion reports were written.

Following customary procedures, copies of the draft PPAR were sent to the relevant
government officials and agencies for review and comment. ODESYPANO expressed its
satisfaction with the report but made no substantive remarks. No other comment was


        This is the Performance Assessment Report prepared by the Operations Evaluation
Department (OED) on the above four projects. The Agricultural Research and Extension
Project was approved in June 1990, and closed in June 1997, one year behind schedule, at
which time US$1.7 million was canceled. The Agricultural Sector Investment Project was
approved in November 1993, and closed in December 2000, eighteen months later than
expected; the undisbursed balance (about US$5 million) was rolled into the second
agricultural sector investment project. The Northwest Mountainous Areas Development
Project was approved in December 1993, and closed in June 2001, one year behind
schedule, with cancellation of US$175,000. The National Rural Finance Project was
approved in May 23, 1995, and closed in June 2001, twenty-one months later than

        The objective of the Agricultural Research and Extension Project was to
improve the institutional framework of these services, principally by creating
coordinating agencies for research and extension, strengthening regional facilities, and
improving the link between research and extension. The project was of modest relevance,
correctly identifying the need to decentralize, but spreading resources too thinly, and
failing to give regional research institutes the budgetary autonomy they needed. The
scope for closer linkage between research and extension was constrained by the lack of
adaptive research, which reflects the academic bias of the research institutes.

        In view of the modest performance in achieving project goals, OED rates outcome
as unsatisfactory. Sustainability is rated unlikely because the vehicles financed by the
project are not being replaced and a government-wide recruitment freeze stops the hire of
laboratory technicians and extension agents. Institutional development impact is rated
modest because the required rationalization of the administrative structure was not
carried out, leaving regional research poles inadequately equipped, experimental stations
run down, and local extension offices abandoned. Bank performance is rated
unsatisfactory because a firmer line should have been taken on administrative
restructuring. Borrower performance is rated unsatisfactory because of substantial delays
during implementation.

        The objective of the Agricultural Sector Investment Project was to support
sector growth by financing a set of indispensable public investments, improving the
policy framework, and strengthening the Ministry of Agriculture and other institutions.
No less than 83 percent of the final project cost was accounted for by hydraulic
investments (irrigation perimeters, hill dams, drinking water supply systems). The project
was highly relevant, building on the recommendations of sector work, particularly with
respect to the need to increase water use efficiency. Physical achievements were
substantial. However, the installed capacity of the perimeters is not being fully used, hill
dams are not viable, and water user groups are not paying tariffs sufficient to cover
maintenance costs: together these factors lower the return on the investments made.

        OED rates outcome as moderately satisfactory, because the full potential of the
hydraulic investments remains unrealized. Sustainability is rated likely, on the
assumption that cost recovery will improve as user groups become stronger. Institutional
development impact is rated substantial: contracting of soil and water conservation works
to private firms, privatization of veterinary services, and strengthening of water user
groups all had a positive impact. Bank performance is rated satisfactory reflecting the
solid analytic underpinnings of the project and the generally adequate nature of
supervision. Borrower performance is rated satisfactory, primarily because of the high
commitment to the project.

         The objective of the Northwest Mountainous Area Development Project was
to improve the well-being of the region’s population and involve the rural population in
efforts to halt degradation of the natural resource base, through a mix of investments
including tree planting and road building. The project was a relevant response to the
acute erosion of watersheds in the Northwest, and made a worthwhile attempt to foster
community-led development while taking appropriate steps to allow for the eventual
transfer of the implementing agency’s functions to regular organs of government. Natural
resource management goals were exceeded and income growth, although somewhat
below target, was still substantial.

        OED rates outcome as satisfactory, primarily because the natural resource
management targets were amply exceeded. Sustainability is rated likely because the
participatory approach has built community ownership, helping to make project benefits
more risk-resilient. Institutional development impact is rated substantial in view of the
progress made with community development plans. Bank performance is rated
satisfactory based on the quality of the participatory method developed and the attention
to monitoring. Borrower performance is rated satisfactory, reflecting government’s
commitment and the efficacy of ODESYPANO, the implementing agency.

        The objective of the National Rural Finance Project was to promote
creditworthy investment in rural areas and strengthen the National Agricultural Bank
(BNA). The project’s design was of negligible relevance because it failed to tackle the
problems—particularly weak loan recovery—that were evident from four previous
agricultural credit recovery projects. Only 26 percent of loans went to small farmers
(compared to the appraisal target of 42 percent)and the recovery rate averaged a mere 39
percent—compared to 63 percent and 54 percent respectively for the Bank-supported
third and fourth agricultural credit projects. A small component intended to explore the
feasibility of village savings and loan associations was dropped.

         OED rates outcome as highly unsatisfactory because the project failed to improve
on the deficient approach and mixed performance of the projects that preceded it.
Sustainability is rated unlikely based primarily on the weak loan recovery rate.
Institutional development impact is rated negligible because, although it introduced a
portfolio monitoring system, the project failed to tackle the central problem of BNA’s
inability to resist government interference in loan rescheduling. Bank performance is
rated highly unsatisfactory because of the design deficiencies (reflected in hurried
preparation). Borrower performance is also rated highly unsatisfactory, reflecting less on

BNA than on government’s failure to allow BNA to function as a solvent commercial

       Experience with these projects confirms four important OED lessons of particular
relevance to future Bank assistance.

         First, a piecemeal approach to reform—trying to effect small changes under the
cover of a number of investment operations—offers little leverage and is therefore
unlikely to work. Piecemeal approaches to reforming land tenure and rural financial
institutions have not worked in Tunisia.

        Second, attempts to strengthen monitoring are futile if the performance indicators
monitored are not linked to the achievement of significant outcomes. Research and
extension indicators, indicative water tariffs, and BNA portfolio indicators—all have
significant drawbacks. For example, although the monitoring of research and extension
has improved most of the indicators measure inputs rather than outcomes.

        Third, decentralization is an important objective, but it may be counterproductive
if resources are spread too thinly. This is a particularly important issue for agricultural
services, the structure of which needs to be rationalized.

        Fourth, where there are multiple projects overlapping in time and by theme, it is
important to assess the overall impact of assistance: where there have been several
interventions which are cumulative in effect it is reasonable for evaluators to have higher
expectations of development effectiveness and to be correspondingly more demanding in
their search for evidence of effectiveness. The upcoming Tunisia Country Assistance
Evaluation will need to rise to this challenge.

                                                             Gregory Ingram
                                                             Operations Evaluation

1.       Background
1.1      This performance assessment of four rural projects is intended to contribute to a
broader, cross-sector evaluation of the Bank’s assistance to Tunisia since 1990.1 Tunisia
occupies a strategic position in the Bank’s regional portfolio of rural projects. Adding up
total lending in 1990–2003, Tunisia accounted for 18 percent of all rural projects in the
Middle East and North Africa Region, and the same proportion of commitments.2 This is
a strikingly large share, considering that Tunisia accounts for only 3 percent of the
region’s rural population. Moreover, Tunisia is a good performer. Between 1991 and
2003, 86 percent of completed rural projects in Tunisia had a satisfactory outcome,
compared to regional average of 63 percent and a Bank-wide average of 65 percent.3

1.2      With assistance from two Bank-supported agriculture structural adjustment loans,
Tunisia has made a series of reforms to the sector since the late 1980s.4 Subsidies for
fertilizer, animal feed, seed, irrigation, and mechanized services have been substantially
reduced. The supply of farm inputs, collection of produce, and the provision of mechanized
plowing and harvesting have been privatized. The role of private extension agents and
veterinarians has expanded. Progress has been slower in liberalizing food marketing, the
state remaining involved in cereals, olive oil, sugar, tea, coffee, tobacco, and wine. Limited
security of tenure, holding fragmentation, and land titling and registration problems are
issues that still await resolution. Water is a key constraint and the importance of an
integrated management of ground and surface sources has been recognized.5

1.3    Three Country Assistance Reviews6 for Tunisia have broadly agreed on the
following priorities for agriculture and rural development:

     •   Increase the competitiveness of agriculture;
     •   Promote sustainable use of water and soil resources;
     •   Strengthen agricultural support services, with privatization where appropriate;
     •   Improve living standards in remote rural areas; and
     •   Pursue dialogue on land titling and tenure issues.

1.5      Three of the projects examined in this report were consistent with these priorities.
The Agricultural Research and Extension Project and the Agriculture Sector Investment
Project addressed competitiveness, efficiency of agricultural services (including
privatization of veterinary services), and better use of scarce water resources. The
Northwest Mountainous Areas Development Project aimed to improve resource use and
raise living standards in the poorest region of Tunisia. On the other hand, the National

1. FY04 Tunisia Country Assistance Evaluation.
2. Business Warehouse data, April 2003.
3. OED ratings database. See also Annex C of this report, Table C1.
4. See Annex A, Agriculture and Rural Development Overview
5. World Bank, Tunisia, Country Assistance Strategy, March 28, 2000, Annex B9a, p. 4.
6. Dated April 1993, June 1996, and March 2000.

Rural Finance Project, with its focus on working through a state-controlled bank, does
not fit well with the Bank’s assistance agenda. Also, there was no sustained dialogue on
land tenure problems, although these issues received some cursory attention under the
sector investment project and the northwest project.

1.6      The Bank’s analytic work suggests that the key challenge is managing the
transition to a smaller, more efficient agriculture sector. The competitiveness study
found that there would be an economy-wide gain from removing farm protection equal to
as much as an additional 1.7 percent of GDP growth per year from 1998 to 2025.7 But the
social and political cost would be substantial. There are some 470,000 farms in Tunisia
out of which 85 percent earn less than US$4,500 per year but employ between 2 and 7
persons per farm. Many of these farmers would be driven out of business.8 Although
labor would ultimately be absorbed elsewhere there would be considerable short-term
pain—stress on urban infrastructure resulting from increased influx of workers from the
countryside, possible increases in illegal migration abroad, straining relations with the
European Union, and a worsening of the livelihoods of aged farmers, too old to adapt.
Many of these stresses would be felt in the Northwest region—the granary of Tunisia—
underlining the need for a program of area development to help smooth the transition.

1.7     The four projects assessed in this report form part of a series (Table C3) that
overlap closely, in time and by theme. This report assesses each of the four projects in its
entirety (see Chapter Six and Annex B). But the discussion in Chapters Two to Five gives
pride of place to four cross-cutting topics that are central to the projects (Table 1). Annex
B provides the complete list of objectives and components for the four projects.

Table 1. Evaluation Design
                                                               Evaluation Topics
Projects                                Agricultural     Rural Water         Area              Rural
(Date Approved to Date Closed)           Services        Management       Development         Finance
Agricultural Research and Extension
Project                                      X
(Jun 1990 to Jun 1997)
Agricultural Sector Investment
Project                                      X                 X
(Nov 1993 to Dec 2000)
Northwest Mountainous Areas
Development Project                          X                 X                 X                X
(Dec 1993 to Jun 2001)
National Rural Finance Project
(May 1995 to Jun 2001)

7. World Bank, Tunisia, Agricultural Competitiveness: A Policy Note on Transition Issues in the Process of
Agricultural Liberalization, July 25, 2001, p. iv.
8. Data from Tunisia Poverty Strategy. The brunt of the adjustment would be borne by small-scale cereal
farmers—cereals being the least competitive sector.

2.       Agricultural Services
2.1     The appraisal report for the Agricultural Research and Extension Project
identified the following weaknesses in the Tunisian program (c.1990):

     •   Research topics selected without considering end users’ needs;
     •   Too many experimental stations—46, many in the same ecological zone;
     •   Poor pay pushing out the most-skilled research staff;
     •   Excessive concentration of staff in Tunis;
     •   Item by item clearance of expenditures required from the Ministry of Agriculture;
     •   Lack of coordination between different extension agencies;
     •   Extension messages not sufficiently adapted to local conditions, out of date;
     •   No feedback from extension to research.

2.2    Thirteen years later—and following a commitment of almost US$40 million 9—
none of these issues have been fully resolved. On the research front, some progress has
been made in four areas; although, in each case, problems remain.

2.3     First, farming and agro-industry groups—the end users of research—have been
given more say in the setting of the research agenda, and greater participation in the
process for reviewing research results. Ten research themes of “national priority” have
been identified, each accompanied by a commission in which end users are represented.
However, the commissions are headed by scientists, not by users; because the scientists
do not speak the language of the users, user attendance at, and participation in,
commission meetings is limited. It was envisaged that users would account for 30 percent
of the membership; but, in practice, only 5–15 percent of participants are users.

2.4     Second, the results of each research project are now presented and discussed in
workshops. The workshops are well attended by users and by extension service staff. But
the framework for evaluating research results (professional peer review, with recourse to
foreign experts where necessary) is poorly developed; and no attempt has yet been made
to measure the impact of research.

2.5      Third, the government has steadily increased the budget for research10 and waived
the government hiring freeze to recruit 10 additional researchers each year: this makes for
a relatively attractive work environment, and greater retention of the best staff. On the
other hand, the freeze still applies to laboratory technicians, which means that
researchers’ time is wasted doing tasks that do not correspond to them.

9. This comprises the actual cost of the Agricultural Research and Extension Project (US$20.9 million) and
the agricultural service component of the second Agricultural Sector Investment Loan (US$7.5 million); plus
the relevant components of the Agricultural Support Services Project (estimated at US$10.0 million). It does
not include the cost of the adjustment operations, which also had some bearing on the evolution of agricultural
support services.
10. Between 1988 and 1996, investment budgets allocated to agricultural research institutes increased by 28
percent in real terms.

2.6     Fourth, the proportion of researchers located in the regions has increased—40
percent now compared to 10 percent in 1990. But the majority of the 198 researchers in
the system are Tunis-based. Also, at the regional level, there has been insufficient attempt
to integrate higher education with research, reducing the scope for developing the critical
mass of intellectual resources needed for programs to flourish.

2.7     The biggest constraint is the administrative structure of research, which has not
evolved enough to permit effective use of resources. There is a cumbersome arrangement
whereby power is divided between three theme-specific institutes (INRAT, INGREF,
IO),11 and one coordinating agency (IRESA). IRESA controls research operating costs
and therefore has an implementing as well as a coordinating role. IRESA is also
responsible for recruitment and procurement, and for the agricultural schools.

2.8     I The research institutes enjoy substantial autonomy. Maintaining these separate
institutes is not cost-effective. Each institute has its own network of research stations and
own headquarters administration. Little progress has been made in rationalizing the
network of experimental stations: out of the 46 stations that existed in 1990, only seven
have either been closed down or given back to the Ministry of Agriculture. Budgets do
not cover basic maintenance, so the stations tend to be rundown; also, scientific work
tends to be squeezed out by covert commercial projects. There have been wrangles over
whether to give the institutes budgetary compensation for the loss of farm income that
would result from the closing of stations.

2.9     Decentralization has been held in check by the failure to grant autonomy to
regional centers: the proposed number and shape of these centers has varied substantially
over the past decade and there has been substantial delay in equipping them. The decision
made in 1992 to move from three regional centers to seven “poles” was hard to justify.
There are now supposed to be seven regional research poles but only two (Beja, Tozeur)
are operational. The design is flawed because the poles have no budgetary autonomy and
the coordinator in charge of each pole (paid by IRESA) has limited power to coordinate
because his researchers are paid by, and respond to the agendas of, the various central

2.10 There is a vision of what the structure for research should be—a vision developed
under the aegis of the Bank-supported Agricultural Support Services Project, which was
approved in 2001. The three central institutes would be dissolved into a single executive
agency. The regional poles would be given greater autonomy. The research commissions
could be made more client-responsive by transferring leadership from the scientists to the
farming and agro-industry groups. An independent scientific panel might be created to
advise on research proposals generated within the commissions; and this panel could also
organize expert review of research findings. The decree fusing the institutes was issued in
January 2003; but it is not clear how long it will take to be implemented. The number of
the regional poles, and their responsibilities, remains up in the air. The future role of
IRESA is undecided: it might become a pure programming and evaluation body, ceding
its procurement and recruitment responsibility to the new single research institute; this

11. There is a fourth institute devoted to animal health (IRVT) but this does little real research.

would also entail its handing over responsibility for agricultural higher education—but to
who? There are still more questions than answers.

2.11 While there has been some progress with research, extension has stagnated. The
Agricultural Research and Extension Project was committed to a training and visit model
that has substantial limitations: it is a costly, blunt instrument, requiring large numbers of
staff but incapable of delivering messages that treat the farm as a diversified business (it
focuses on commodities in isolation).12 In principle, the T&V model fits the political
structure like a glove. Tunisia’s regional administration comprises 24 governorates, each
divided into delegations, which are further divided into wards (imadas). The extension
service provides a coordinator who sits in the Agriculture Commission (CRDA) of each
province, with the line of command running down to offices located, first, at the
delegation level (the CTVs), then at the imada level (the CRAs).

2.12 In practice, the extension service is a shell. The extension coordinator cannot
coordinate because the real power lies with the head of the CRDA (who can assign
extension staff to other duties). Extension agents spend only half of their time “on task”
(Table A4). There are insufficient staff to occupy all the posts on the organizational
pyramid. CTVs are supposed to be headed by engineers but are often staffed with less
qualified people. Many of the 768 local extension units (CRAs) are not staffed: in Beja,
32 out of 49 are functioning, in Sfax, 25 out of 50.

2.13 Under the auspices of the Agriculture Research and Extension Project, a central
coordinating agency for extension and training (AVFA) was set up in 1991. AVFA has a
monitoring system that annually tracks the work of each extension officer. It measures
inputs (numbers of visits, information days, demonstration plots) but not outcomes.

2.14 Some progress has been made in improving research and extension linkage. CRDAs
and extension services participate in the research committees. Farmers and extension agents
attend information days at research stations. There is some ad hoc problem solving by
researchers in response to problems identified by CRDAs. But the scope for closer linkage is
constrained by the lack of adaptive research, which in turn reflects the academic bias of the
research institutes (where staff evaluation is based primarily on publications’ record), and
the continuing concentration of researchers in Tunis.

2.15 AVFA says that the technical summaries produced at the completion of each
research project are hard to operationalize. IRESA says that it is not its job to develop
extension messages. Developing new messages is not treated as a collaborative exercise.

  The Region comments “The needs of the agricultural producers are evolving, given the changing
environment (trade liberalization, the need to improve agricultural competitiveness and product quality).
They need more specialized advice or advice on overall farm management”.

3.       Rural Water Management
3.1     The 1991 expenditure review for the agriculture sector emphasized the need to
raise the efficiency of water use—“This must be the government’s number one priority in
the agricultural sector.”13 The hydraulic focus of the Agricultural Sector Investment
Project—83 percent of final cost—was potentially consistent with this recommendation;
but it has not made a significant contribution to raising efficiency.

3.2      The project’s physical achievements were substantial.14 The construction and
rehabilitation of irrigated perimeters was the single largest cost item in the project. A
recent impact evaluation examined a representative sample of 13 perimeters (3,485
hectares).15 The study finds that on none of the perimeters is the irrigated infrastructure
being fully used (despite the 1999–2002 drought), and that there is little sign of crop
intensification. Dividing cropped area by irrigated area yields on average a figure of 1.23,
lower than what the feasibility studies projected (Table C7).16 The report observes that
“irrigated” farmers are still behaving as if they were rain-dependent. Farmers are tending
to favor low-margin crops that need less water (cereal and forage), rather than increasing
the share of land in high-margin fruit and vegetables. In 9 of the 22 farm models studied,
over one-half of the irrigated area was in cereals and forage. This may partly reflect
deficiencies in the quantity and quality of water delivered. Of the 159 farmers
interviewed for the study, only 30 percent said they were satisfied with the access to
irrigation water. Moreover, the limited use of complementary inputs is depressing crop

3.3    One of the perimeters included in the study (Bougossa II) was visited during the
mission. The perimeter has an equipped irrigable surface of 80 hectares; but, in 2002,
only 30 hectares were actually irrigated. Combining the cost data supplied to the mission
with the data from the farm models in the impact study shows that, at the present rate of

13. Agricultural Expenditure Review, May 20, 1991, p. iii.
14. See Annex B of this report.
15. CNEA, Etude d’evaluation d’impact socio-economique des perimeters irrigues (draft report, February
2003). The study has several limitations. It compared the situation before and after the project (rather than
“with” and “without”). Also, there was no cost-benefit analysis: by themselves, the data on incremental
crop revenues and employment shed little light in the absence of other data on how much the irrigation
works cost to build, operate and maintain.
16. In 2000/01, the average rate of intensification for all perimeters—not just the sample in the impact
study—was 91 percent (Ibid., p. 8).
   The Region comments: “The mise en valeur question is serious. As the recent CNEA study points out,
even irrigated farmers behave like rainfed farmers with an overall cropping intensity only at 90 percent.
Again, this is highly variable but the fact remains that this valorization aspect has not been given the
attention it deserves”. A consultant who worked on the implementation completion report for the follow-up
to the Agricultural Sector Investment Project notes that although returns to the irrigation investment are
below appraisal projections, revenues are nevertheless significantly above the pre-project level and it is still
too early to judge the ultimate impact of the investment. He made a separate estimate of the level of
intensification for 6 of the 13 perimeters referred to in the CNEA study (Table C7), finding that average
share of the irrigated area in cereals was 21 percent (compared to CNEA’s finding of 35 percent). (Personal
communication, Christian Lauwers).

underutilization, the financial rate of return is generally under the 12 percent opportunity
cost of capital (Table C8). Bougossa II does not appear to be an exceptionally poor
performer: 15 of the 32 farm models nationwide have per hectare returns lower than the
lowest return model at Bougossa II (Table C7). Also, given the high level of input
subsidies in Tunisian agriculture, the economic rate of return would be substantially
lower than the financial rate of return. This suggests that, without fuller use of the
installed capacity and crop intensification, the investment in the perimeters would not be
economically justified.18

3.4      Similar doubts exist about the efficiency of investments in hill dams (14 percent of
actual project cost). At completion, 14 of the 15 dams were inspected, and 9 were found not
to have realized their potential. The completion report states that there were sporadic efforts
by farmers to build irrigation works around the dams (this was not included in the project)
but with little evidence that the expected agricultural benefits were being realized. The
report also notes that, “sometimes,” development costs per hectare were high, cropping
intensities were low, and the choice of dam sites paid insufficient attention to the potential
impact on agricultural growth. A visit to the Massila Thibar dam in Jendouba revealed that
of the irrigable surface of 180 hectares only 84 hectares had been equipped with spray
irrigation. Dams of this type tend to silt up rapidly and one farmer suggested that about
one-third of Massila Thibar’s capacity had already been lost. All in all, the investment was
probably not worth it. The lengthy process of expropriating and compensating farmers
whose lands would be flooded—306 farmers were affected although only one had to
relocate—required close Bank supervision, pushing up administrative costs. Significantly
perhaps, hill dams did not figure among the items to be financed in the follow-on sector
investment loan.19

3.5     The project supported the creation and training of water user associations, both
for drinking water supply (185 groups) and for the irrigation perimeters (73 groups). Each
of the Regional Agricultural Development Commissariats (CRDAs) was equipped with a
mobile multidisciplinary team that worked at strengthening the technical, administrative,
and financial capacities of water user groups. The user groups still depend heavily on
CRDAs for guidance, and for major repairs; and the CRDAs’ supervision capacity is

   At Bougossa II, taking the perimeter as a whole, profitability is less than the average for other perimeters
owing to the relatively high proportion of land in cereals (25 percent, compared to the average of 9
percent), and forage crops (25 percent compared to 12 percent). This helps to explain the low financial rate
of return for this perimeter. On the other hand, the assumption in Table C8 that benefits remain constant
from year 4 to year 20 may too pessimistic. (Lauwers, ibid. )
   The Region comments that there was an unrealistic assumption that irrigation works (not financed by the
project) would be built by beneficiaries; these investments have not so far materialized so the benefit
stream is lower than expected. On the other hand, the economic analysis failed to take into account other
benefits—including improved access to potable water, protection of downstream areas against flooding,
groundwater recharge—focusing exclusively on the direct impact on crop and livestock production. The
economic analysis treated each dam in isolation rather than considering what role it played in the broader
catchment area or small river basin of which it was a part. These considerations were taken into account in
design of the follow-on project—in which hill dams initially figured prominently; but the component was
dropped before the project was approved owing to the anticipated difficulty in meeting legal requirements
for resettlement of the affected population.

severely overextended.20 User groups in charge of the drinking water systems have been
less effective at recovering costs than those in charge of the irrigation perimeters; and
even the latter are far from being financially autonomous.

3.6     A key policy objective was to increase water tariffs to ensure adequate coverage
of operations and maintenance costs. The project set a target of achieving an annual 9
percent real increase in the tariffs applied to public schemes (those managed by CRDA).
This was achieved in 1994–98. The average increase rose to 14 percent in 1998/99, but
then fell back: to 3 percent in 1999/00, 6 percent in 2000/01 and then with no increase in
2001/02 (a concession to the continuing drought).

3.7     In 1991–2001, the actual price paid for water was less than the indicative tariff.
The indicative tariff published each year for perimeters managed by the CRDAs averaged
TD 0.103/m3 (US$ 0.07/m3) in 2001. This same year, the price paid in the 13 perimeters
included in the impact study—DT 0.070/m3—was one-third less than the indicative price
(Table C7). Data from another source show that, in 2002, in Jendouba, for 11 perimeters
managed by user groups, the tariff averaged only TD 0.061/m3.21 These findings suggest
that, despite claims to the contrary,22 there is insufficient provision for maintenance. A
similar problem exists for drinking water supply systems: only 20 percent of the 185 user
groups created under the project charge a tariff high enough to cover operations and
maintenance costs; one-third of them recover less than 50 percent of operating costs.23

4.       Area Development
4.1     The project assessed in this evaluation is the third of four Bank-supported
Northwest development projects. The first project sought to persuade rural communities
to intensify livestock raising by improving pasture on the middle slopes. This would free
the upper slopes for reforestation, while (intensified) cropping would be limited to the

20. Etude d’Evaluation d’Impact Socio-Economique des Perimetres Irrigues, February 2003 draft, p. 44.
The Region agrees that the extension program has not kept pace with the rapid growth in the number of
water-user groups, noting that the number of groups per CRDA extension agent rose from 20 in 1997 to 30
now. Also, ancillary service provision is often weak. The part of the Agriculture Ministry responsible for
irrigation (DGGR) “needs to work on broadening inter-ministerial support for the GICs [user groups]. In
governorates where this has happened, especially at the delegue level, the program moves forward”.
21. Rapport de Suivi, Annexes GIC d’Irrigation, Jendouba, 2002 (Draft supplied to mission).
22. « La mise en oeuvre de cette politique d’augmentation des tarifs de l’eau a permis d’améliorer de
manière significative les taux de recouvrement des coûts de fonctionnement et d’entretien des périmètres
publiques irrigues,” Pre-evaluation, Projet d’Investissement du Secteur de l’Eau, Note Technique:
Tarification de l’Eau d’Irrigation (Draft supplied to mission in March 2003. p. 2)
  The Region notes that recovery of operations and maintenance costs varies greatly between user groups
in different regions. A study of 384 groups nation-wide found that 18 percent recovered less than half of the
costs, 32 percent recovered between 51 and 80 percent, and 50 percent recovered between 80 and 100
percent. The average cost recovery rate is 82 percent. The Region describes this as “encouraging despite
the fact that this does not include the cost of large repairs or depreciation which is still borne by the state”.

lower slopes. Classical soil conservation measures were installed, with limited
participation of the local population.

4.2     OED’s evaluation report for this first project observes that the appraisal targets for
infrastructure on public land (schools, drinking-water systems, and access roads) were
exceeded; the project was less successful at installing erosion control works on private
land (farmers would not accept tree planting but accepted stone contour terraces on
uncultivated land because de-stoning of the adjacent land left it cultivable); and it failed
to substantially change farming practices (farmers were not willing to stop cultivating
upper slopes and cropping was not much intensified on the lower slopes).

4.3     The implementing agency, ODESYPANO is one of a group of agencies (Offices)
set up to deal with specialized problems that are often deemed to be temporary in nature
and limited to a particular region. The Offices are not part of the regular apparatus of
government, although the length of their mandate is unclear. When ODESYPANO was
created in 1981—the Bank’s support for it prevailing over the borrower’s initial
resistance—its remit was to protect the fragile ecosystems of the Northwest region while
promoting the development of rural infrastructure. OED’s first impression of
ODESYPANO was that it was good at engineering, but ineffective as an extension
agency because it was too centralized and not responsive enough to villagers’ needs.

4.4     The physical achievements of the Northwest Mountainous Areas Development
Project were impressive.24 To reduce erosion, a key aim was to substitute temporary crops
with permanent crops and improved pasture. As a share of the cultivable area, cereals
went down from 48 percent in the 1996 baseline survey to 40 percent in the 2000 survey;
tree crops increased from 10 percent to 22 percent—planting of olive trees being
particularly widespread.

4.5     Income growth was to be promoted through crop and livestock development, and
by off-farm activities, supported by improved financial intermediation. Crop incomes
reached 56 percent of the target for farmers with holdings under 5 hectares; and 83
percent of the target for farmers with 5 hectares or more.25 Crop income in 2000 was 47
percent higher than the 1996 baseline, in the case of small farmers, and 50 percent higher
for larger farmers. This partly reflects growth in yields: between 1996 and 2000, wheat
output per hectare increased by 18 percent for small farmers and 50 percent for larger
farmers. Milk and meat production also increased substantially.26 The impact on farm
employment was substantial. The number of days spent working on-farm increased from
the baseline of 136 per year to 285 per year at project end.

24. See Annex B. Targets quoted are those adopted after appraisal, in line with Quality Assurance Group
efforts to retrofit key indicators for projects effective after 1994.
25. The targets were respectively TD2,900 per year and TD5,700 per year. The impact on incomes is
derived from a 1996 baseline survey and follow-up surveys in 1998 and 2000, involving 259 farm families.
26. This might seem surprising given all the talk in Tunisia about the severe impact of drought in 1999–2001.
However, for each of the three provinces in the ODESYPANO region (Beja, Jendouba, Le Kef), annual
rainfall was higher in 2000 than it was in 1996 (Annuaire des Statistiques Agricoles, 2001, Table 1.3).

4.6     At appraisal and at completion, an economic rate of return was estimated based on
models for small, medium, and large farms representative of the three geographic zones
covered by the project. The cost stream included the agricultural production component,
the watershed and rangeland management component (excluding the soil conservation
aspects) and the cost of rural roads. The benefit stream covered incremental crop output,
including fodder, and the savings in vehicle depreciation costs as a result of road works.
The economic rate of return was 14 percent at appraisal and 16.3 percent at completion.
The assumptions seem plausible, and the results consistent with the findings of the
beneficiary surveys. OED did not re-estimate the rate of return. This positive outcome is
only somewhat offset by the higher-than-expected unit costs of schools and health posts:
the appraisal underestimated the charges that contractors would make for operating in
difficult-to-access areas.

4.7     Strictly speaking, this is not an example of demand-driven development because
ODESYPANO’s mandate for natural resource conservation is uppermost in shaping the
village development agenda. Also, the community contribution (finance, labor, materials)
is low. There is a long way to go. Community organization is still fragile and will
continue to require capacity building to ensure that benefits are sustained. (Box 1).

Box 1. Village Development at Traia, Bizerte
The Development Committee was created in 1997 and the first community plan was prepared in
1998. There are 7 members (including 1 woman), representing 106 families, divided between
three hamlets. At first, the villagers were difficult to mobilize. Agriculture development is
difficult owing to the extreme fragmentation of holdings; so far, the government’s program to
consolidate holdings has not touched hillside communities. Improved cattle have been introduced
and olive trees planted, with respectively 50 percent and 85 percent of the investment cost
subsidized by the government. The Development Committee hopes to acquire the legal status of
an Agricultural Development Association (GDA), which would enable it to receive and manage
funds. Significantly perhaps, when the mission visited this village, the presentation of the
community development plan was made by ODESYPANO officials, not by the committee. The
committee has taken initiatives on its own—for example, hiring the machinery needed to open a
road—but they don’t yet have a treasury, there is no savings initiative, and the Development
Committee has little outreach to government agencies other than ODESYPANO.

4.8     The project aimed to establish village Development Committees. The aim was to
create 230 active committees, but only 162 were established. These committees required
careful supervision. Rather than spread its resources thinly, the project team chose to
work more intensively with a smaller number of committees. ODESYPANO argues
plausibly that the number of committees formed is less important than the progress made
with community development plans—now completed for one-third of committees (Table
C6).27 Earlier attempts to finance investments before development plans were ready
(using the system of Integrated Action Contracts) resulted in less viable projects—
because technical assistance and training was lighter—and this approach was dropped.

27. This table refers to the whole ODESYPANO region, not the sub-region covered by the Bank-supported
project (in which 162 committees were created).

4.9     Community capacity building is a slow process in all countries; in the highly
centralized environment of Tunisia, what ODESYPANO has achieved in the space of six
years is substantial.28 Prospects are good. A law passed in 1999 allows the committees to
evolve into Agricultural Development Associations (GDA), empowered to receive and
manage public funds. A further decree, issued in December 2002, stipulates the ceiling
(equivalent to roughly US$50,000) that groups of this nature are able to manage.

4.10 The follow-on project is anticipating the eventual handover of ODESYPANO’s
responsibilities, which makes benefits more likely to be sustained. Provision is made for the
project’s schools and health posts to be operated and maintained by the relevant ministries.
However, the key partners—the Regional Agricultural Development Commissariats
(CRDAs)—are still poorly placed for the transition. They do not use a participatory
approach and their extension agents are not as close to farmers as ODESYPANO. Also, they
have not previously had any responsibility for the hillsides (focusing on the higher-potential
bottomlands). Transfer is not expected to begin before 2005; but this date will probably be
pushed back. A major re-zoning exercise is entailed because ODESYPANO’s
administrative subdivisions do not correspond to those used by the CRDAs.

4.11 Road maintenance has been a problem. Roads were not built to specifications of
the Equipment Ministry, which therefore resists a blanket assumption of maintenance
responsibilities. A study is needed to classify all the roads built by ODESYPANO since
1981 and the corresponding level of maintenance that is deemed appropriate.
Responsibility for feeder road maintenance could possibly be assumed by the Regional
Agricultural Development Commissions (CRDAs).

5.       Rural Finance
5.1    The five Bank-supported agricultural credit projects all worked through the
National Agricultural Bank (BNA)—the largest commercial bank in the country with a
network of 13 regional offices and 126 branches. Past evaluations of this institution—or,
more appropriately, the broader institutional framework of which it is a part—have been
remarkably forgiving. A profound rethinking of the approach to rural finance is long

5.2     In 1993, an OED mission concluded that, despite the weak loan recovery rates
achieved by the third and fourth agricultural credit projects, their outcome was
satisfactory. The mission’s recommendations for future operations included:

     •   Raise BNA’s interest rates to market levels, thus increasing its incentive to lend to
         farmers, and to devote more staff time to loan recovery;
     •   Encourage stricter loan appraisal by BNA by discouraging government from
         rescheduling or forgiving BNA’s bad loans;
  The project was able to capitalize on lessons learned from a previous project supported by German
technical assistance; with this experience behind them ODESYPANO staff were well placed to tackle the
new project (Personal communication, Christian Lauwers).

     •   Increase the authority of branch offices in approving and administering loans;
     •   Harmonize loan terms across all BNA programs;
     •   Establish a private sector-operated calamity fund, with crop insurance paid by
         sub-borrowers, complemented by funds from the government budget;
     •   Prepare a specific balance sheet for the agricultural portfolio—separating it from
         commercial and industrial loans—making transparent the losses that BNA makes
         on agricultural credit and the extent to which these are absorbed by government;
     •   Increase BNA’s outreach to viable smallholders incorrectly regarded by BNA as
     •   Address the problems posed by insecure land tenure and fragmented farm
         holdings; and
     •   Reorient rural transfers by replacing subsidies granted through the credit system
         by transfers through other channels (e.g., rural infrastructure spending).

5.3     OED’s report welcomed the proposals for the forthcoming National Rural Finance
Project, commending the planned reorientation of BNA’s training program toward small-
farmer lending, diversification into non-farm rural finance, banking services for women,
and involvement of new intermediaries.

5.4     However, the project did not come to grips with these challenges. There was no
clear commitment by government to letting BNA assume responsibility for its
rescheduling decisions, and for several years after the project was approved, BNA did not
have the systems to allow it to distinguish between loans granted before and after the
project, making it hard to track improvements in recovery performance.

5.5     The project failed to meet its objective of promoting creditworthy rural
investment, first, because the share of loans directed to rural enterprise was lower than
expected and, second, because the loan appraisal process was insufficiently rigorous,
resulting in low repayment rates.

5.6    At appraisal it was estimated that 60 percent of project funds would be on-lent to
farmers (42 percent to small and medium-scale farmers). By closing, 51 percent of Bank
disbursements had been used to finance loans to farmers; and only 26 percent to small
farmers. Agro-industries received the largest proportion of financing, absorbing 39
percent of funds disbursed from the Bank loan (compared to the 15 percent share
envisaged at appraisal).

5.7     During implementation the repayment rate of sub-loans varied between 26 percent
and 48 percent, averaging 39 percent by project close. BNA’s overall repayment rate (i.e.,
taking into account all sub-loans, not just those financed under the Bank project) was 74
percent. For the two preceding Bank-supported projects involving BNA the recovery
rates averaged 63 percent and 54 percent.29 Many farmers regard loans as subsidies and
the government’s willingness to reschedule and forgive reinforces this attitude. Recovery
rates were substantially higher for loans to commerce and industry (averaging 95 percent
  Rates for the Third and Fourth Agricultural Credit Projects, respectively (see OED Performance Audit
Report, No. 11977, June 11, 1993, p. 34).

in 2000/01) than for loans to agriculture (averaging 45 percent). BNA acknowledges that
it aims to offset its losses on the agriculture portfolio by lending to non-rural enterprises.

5.8      The completion report states that one of the project’s principal successes was the
strengthening of BNA’s financial position. The mission updated a key table in the
completion report, finding some slippage in relation to targets set in 1999 (Table C10).
The proportion of loans overdue rose between 2001 and 2002 and there has been no
improvement in the share of agricultural loans at risk (around two-thirds). In truth, this
performance monitoring is no more than an academic exercise as long as BNA is unable
to resist government interference in rescheduling and debt forgiveness.

5.9     The project made a very modest contribution to improving data processing and
management information systems but spending was limited to US$111,000. The mission
visited branch offices in Beja and Sousse, and observed that they are poorly equipped to
track non-performing loans.

5.10 The project sponsored a study of small-farm finance.30 It estimates that 22 percent
of the total of 409,000 small farmers are “bankable.” But the study found that small-scale
farmers make little use of credit (only 35 percent of those interviewed had taken loans).
Second, 73 percent of credit users had difficulty in repaying the loans received. Third,
input suppliers play a key role in providing credit and invariably are repaid (consistent
with findings in many other countries). Finally, taking all sources of credit together,
repayment rates are no higher for irrigated farmers than for rain-fed farmers: this raises
doubts about the convention of attributing defaults primarily to drought (which typically
affects rain-fed farmers more than irrigated farmers).31

5.11 The mission took a closer look at the relationship between drought and loan
recovery levels. This is an important issue because drought is frequently invoked to explain
the weak rate of loan recovery; and the recent “four years of drought” (1999–2002)
prompted government to order a major loan rescheduling in August 2002. Twelve sites
scattered throughout the country were selected, chosen because they each have a BNA
branch office and a meteorological station, with a complete set of data for rainfall and the
volume of loans recovered between 1994 and 2001. It might be expected that there would
be a tighter correlation in rain-dependent areas than in irrigated areas; but there is no
evidence of this (Table C9). With one exception (Jendouba), there is very little correlation
between rainfall and recovery. Jendouba is indeed a rain-dependent, cereal growing area,
reportedly badly affected by the recent drought. Therefore, the mission took an even closer
look at this province, extending the enquiry to all branch offices. Of the six offices, only
two of them have coefficients of 0.80 or higher; for the other four offices there is no
evidence of a significant positive association between rainfall and loan recovery rates.
Drought is a long way from being a sufficient explanation of BNA’s poor record.

30. This was substituted in October 1999 for the planned feasibility study of village-level group lending.
31. SCET, Etude de la Petite Agriculture a Caractere Familial et Social: Note de synthese des travaux de
Phase I et de Phase II, October 2001, p. 8.

5.12 The obstacles to developing new forms of rural financial intermediation are
considerable. First, the government has not shown itself to be interested. This helps to
explain why the planned feasibility study of village savings and loan associations was
shelved. In 1999, when the decision to drop the study was made, the government agreed
with the Bank that BNA would sponsor a pilot to provide financing to agricultural service
cooperatives. The mission found that this initiative has also failed to prosper.

5.13 There is a lack of political will to remove legal obstacles to mutualiste finance: a
1959 law restricts village groups from forming saving and loan associations. This
explains why, in the Northwest Mountainous Areas Development Project, a plan to
establish village (douar) credit unions was dropped. The alternative that was developed—
extending subsidized credit to rural households through the Banque Tunisienne de
Solidarite—also met with little success. Nongovernmental organizations were used to
administer loans, but recovery rates were poor (70 percent at best; and reportedly as low
as 30 percent in Le Kef), partly because loan proposals were not rigorously evaluated.
Despite these setbacks, there have been some promising experiments—under
ODESYPANO’s jurisdiction, GTZ successfully piloted a group-lending scheme in Nefza.
But there is no prospect of scaling up until the law is revised.

6.     Ratings


6.1     The outcome of the Agriculture Research and Extension Project is rated
unsatisfactory. The project had only modest relevance (Table 2) because both the initial
design and the modifications made during implementation worked against realizing the
objectives of rationalization and decentralization. Efficacy was also modest because the
key goals—improved programming, stronger regional facilities, and better linkage
between research and extension—were only partially attained. The project did not include
any explicit efficiency indicators, but the evidence (Table C4) is not encouraging. The
administrative structure inhibits efficient resource use.

Table 2. How the Outcome Rating is Derived
                      Relevance/1           Efficacy/1            Efficiency/1          Outcome/1
Agricultural          Modest                Modest                Modest                Unsatisfactory
Research and
Extension Project
Agricultural Sector   High                  Substantial           Modest                Moderately
Investment Project                                                                      Satisfactory
Northwest             Substantial           Substantial           Substantial           Satisfactory
Mountainous Areas
National Rural        Negligible            Modest                Negligible            Highly
Finance Project                                                                         Unsatisfactory
/1 See frontispiece for explanation of these OED concepts.

6.2     The outcome of the Agriculture Sector Investment Project was rated
satisfactory in the completion report; but this presupposes that the project “makes
effective use of resources (10 percent rate of return or better where it can be
estimated).”32 The overall project rating will be driven primarily by the performance of
the hydraulic investments: the share of water-related investments in total project costs
rose from 71 percent to 83 percent between appraisal and completion. So far, it is not
clear that a satisfactory economic rate of return is being achieved. On the other hand, the
project’s objectives were highly relevant and they were substantially achieved. This
points to an outcome rating of moderately satisfactory.

6.3     The outcome of the Northwest Mountainous Areas Development Project is
rated satisfactory. When OED reviewed the completion report, it lowered the outcome
rating from satisfactory to moderately satisfactory, arguing that farm incomes and user
capacity were not enhanced as much as expected. Based on findings in the field, this
assessment finds that OED’s initial verdict was too harsh. Fewer village development
councils were created than expected, but what counts is the progress with community
development planning. More important, the natural resource management component
(accounting for 40 percent of project costs) amply exceeded its targets. The rate of return
for representative farm models was over the 10 percent threshold.

6.4      The outcome of the National Rural Finance Project is rated highly unsatisfactory,
a slight downgrade from completion report’s rating of unsatisfactory. The rating is driven
by the credit line component, which accounted for 98 percent of the expected project cost.33
Many loans were made, but fewer than expected to small-scale farmers (modest efficacy),
and recovery was low (negligible efficiency). Given the failure to address flaws identified
by previous projects, the relevance of the project was negligible to begin with; and this
diminished further in the course of implementation as the piloting of village group lending
was dropped, and as large-scale agro-industry’s share of the loan portfolio increased.

32. Operations Evaluation Department, Guidelines and Criteria for OED Project Evaluations, July 1, 2000.
(See Section 6, Outcome).
33. There is no data available on final project costs (expected costs were US$420 million).

Sustainability and Institutional Development

6.5     The sustainability of benefits generated by the Agriculture Research and
Extension Project is rated unlikely. The freeze on recruitment—which bears on contract
as well as permanent staff—has hit research less severely than extension; but the ban on
hiring laboratory technicians compromises the research program. The viability of the
extension service has been reduced by the failure to offset depreciation of the vehicle
fleet and the failure to replace departing extension staff.34

6.6     Institutional development impact is rated modest, in view of the limited progress
with efficient decentralization of research, and the slowness to address the need to
rationalize the cumbersome apparatus of agricultural services.

6.7     The Agricultural Sector Investment Project is given a sustainability rating of
likely. With respect to the hydraulic investments—where the burden of proof lies given
their heavy share of total project cost—although the evidence suggests that the installed
capacity is underused and costs are not being fully recovered, it is to be hoped that the
situation will improve as user groups grow stronger.

6.8     The project’s institutional development impact is rated substantial, primarily
reflecting the investment made in creating and training water user groups. Also, the
project encouraged private sector development by ensuring that all soil and water
conservation works were contracted to private firms; and by its support to the
privatization of veterinary services.

6.9     Sustainability is rated likely for the Northwest Mountainous Areas
Development Project. The participatory approach promoted by the project is now tried
and tested, with community ownership helping to make benefits more risk-resilient. In
particular, the rapid take-up of perennials by farmers, and their willingness to take other
soil conserving measures (e.g., planting hedges) shows that they are committed. One
caveat is the failure to reach agreement on which agency will be responsible for
maintaining the roads built by the project.

6.10 Institutional development impact is rated substantial. The project aimed to
enhance the capacity of ODESYPANO and of the community development committees.
In the longer view, the second of these efforts is the more important and the project team
is rightly emphasizing the process of community development planning. The legal
framework is evolving in a manner that will enhance the project’s institutional

    The Region disputes the rating of sustainability as unlikely: “OED bases its ratings on the fact that
extension agents are ageing and are not replaced. This is a relatively crude and unsatisfactory measure of
sustainability. As mentioned above, the information and communication needs of farmers have evolved.
There will be a decreasing need for extension through CRDAs which should no longer be the only service
provider. Indeed it may be a blessing in disguise that extension agents are retiring: it allows the Ministry
to rethink its system”.

development impact. The contracting of soil and water conservation works helped to
build private sector capacity.

6.11 For the National Rural Finance Project, the mission endorses the completion
report’s rating of sustainability as unlikely, based primarily on the weak loan recovery
rate. Low recovery is partly the result of a lack of rigor in appraising loan proposals,
leading to the financing of projects with limited economic viability. Much of the lending
that was achieved would have had a positive short-term impact on user incomes, without
necessarily promoting a long-term improvement.

6.12 Institutional development impact is rated negligible. The completion report rated
institutional development as substantial based primarily on the improvement in BNA’s
overall financial condition. But progress in relation to the benchmarks established in
1999 has been limited (Table C10). The key failing is the government’s refusal to grant
BNA the autonomy needed to manage its portfolio efficiently. Also, an opportunity was
missed to experiment with alternative approaches to rural finance, leading to no
improvement in the effectiveness with which resources are mobilized.

Bank and Borrower Performance

6.13 For the Agricultural Research and Extension Project, performance by each party
is rated unsatisfactory. The completion report indicates that Bank staff did not fully
appreciate the difficulties involved in implementing some project components, especially in
the case of research, where complex changes were envisaged. This seems consistent with a
more general failure to grapple with the challenges of rationalization and decentralization.
The completion report also notes that there was care to avoid introducing pure T&V, which
was thought inappropriate for Tunisia. The marginal cost of the structure that was created
may have been low—because it was grafted on to the existing CRDA system—but the
whole structure needed reformulating because it spread resources too thinly. Furthermore,
the Bank should not have allowed the original plan to create three regional research centers
to be replaced by the proposed creation of seven separate poles. Also, a firmer line could
have been taken over the rationalization of the experimental stations.

6.14 Borrower performance is rated unsatisfactory in the completion report. Although
the government showed commitment during preparation, subsequently there was much
foot dragging on decentralization and rationalization, calling into question the sincerity of
the government’s commitment

6.15 For the Agriculture Sector Investment Project, OED endorses the completion
report’s rating of Bank and borrower performance as satisfactory. Project preparation was
satisfactory, building on a robust policy dialogue with government and based on solid
analytic work. Supervision was also satisfactory, with regular twice-yearly missions staffed
with a wide range of technical expertise—which helps to explain why supervision costs
were double the regional average (Table C11). The primary shortfall seems to have been a
delay in handling hill dam resettlement issues, but overall the Bank provided active and

competent supervision and maintained a sound working relationship with the borrower

6.16 The borrower was highly committed to the project, from preparation through
closing. There were no delays or shortfalls in counterpart funding. The project was well
staffed, with especially strong support from the General Directorate of Finance and
Incentives, the department most directly responsible for implementation. The completion
report notes, however, that coordination with the Regional Agricultural Development
Committees and the entities in charge of the fisheries and livestock subcomponents was

6.17 With respect to the Northwest Mountainous Areas Development Project, OED
also endorses the completion report’s rating of Bank and borrower performance as
satisfactory. Project preparation was satisfactory. The participatory method that was
developed for working with communities improved significantly on the approach taken by
the earlier projects. Supervision was also satisfactory—although supervision costs were
almost double the regional average (Table C11), partly reflecting the need for a
multidisciplinary team. The project team took special care with the baseline survey and two
follow-up surveys, greatly facilitating the evaluation of project results. The emphasis placed
on improving the quality of the Community Development Plans was highly appropriate.

6.18 Borrower performance was satisfactory during both preparation and
implementation. With respect to government performance, the only shortfall was the lack
of commitment to creating a legal basis for the village credit unions. Otherwise,
government strongly supported the project approach, justifying an overall rating of
satisfactory. The ICR rated the performance of the implementing agency, ODESYPANO,
as highly satisfactory, a verdict that, in light of mission findings, OED fully endorses.
The team was innovative and flexible and monitored project progress very carefully,
producing a series of high-quality supervision reports.

6.19 Bank and borrower performance is rated highly unsatisfactory for the National
Rural Finance Project, a somewhat severer verdict than that in the completion report
(“unsatisfactory”). This fifth project was largely a facsimile of the four that had preceded
it, and its design showed little sign that lessons from past experience had been
incorporated. The time elapsed between identification and Board approval was extremely
short (six months). More time was needed to work out the action plan for BNA
restructuring, and to explore new approaches to rural finance. From 1995 to 1999 (i.e.,
for most of the period of project implementation), the completion report provides
evidence that supervision was insufficiently rigorous.

6.20 On the borrower’s side, the shortcomings lie primarily with government, not with
the implementing agency (BNA), because the latter’s hands have been tied by the former.
For example, the completion report states that there was no clear commitment from the
government that BNA would be allowed to make its own rescheduling decisions.

7.       Findings and Lessons

Findings for Tunisia

7.1      Raising agricultural efficiency will not be possible without an institutional
framework that is both lighter and more decentralized than at present. The administrative
structure for agriculture is overly elaborate: the organization chart has too many cells in
relation to the number of qualified personnel, a circumstance aggravated by the
government-wide hiring freeze. Agricultural research remains heavily centralized and
insufficiently responsive to user needs. Agricultural extension is largely decentralized;
but its staff are aging, under resourced and poorly coordinated. Linkage between research
and extension remains weak.

7.2     A key aspect of decentralization is helping local organizations—water user
groups and village development committees—assume responsibility for their
development. Community-driven development is a new concept in Tunisia that needs
time to take root. The recent northwest development projects are pointing the way,
organizing communities to draw up their own development plans. On the other hand, the
deeply ingrained tradition of government subsidy works against the principle of
communal self-reliance and cost recovery; this is particularly evident in the weak
performance of water user groups.

7.3     The government has been slow to ease the (interconnected) constraints of poorly
performing land markets and weak rural financial intermediation. In the northwest, 72
percent of farm holdings under 10 hectares are untitled; the same is true for 50 percent of
holdings over 100 hectares.35 Only in the past year or so has government begun to tackle
this problem.36 Another problem concerns the lack of political will to remove the legal
impediments to group-lending schemes, which use peer pressure rather than land as a
loan guarantee. The authorities remains too heavily committed to a lending institution
(the National Agricultural Bank) that is unable to serve the needs of most of the rural
population, and whose solvency is compromised by the government’s policy of debt

7.4     Tunisia plays a major role in keeping MNA’s rural portfolio afloat, which poses a
risk as much as an opportunity. From the Bank’s perspective, Tunisia has the attraction
of being both a regular client and an above-average performer—qualities that are to be
appreciated in a region where aggregate lending is small and highly volatile, and
performance is below par. Relations between the Bank and Tunisia are highly cordial; but
the Bank’s influence is limited given Tunisia’s easy access to cheaper, less conditional

35. Project Appraisal Document, Northwest Mountainous and Forestry Areas Development Project,
September 26, 2002, p. 16.
   The Region comments: “Regarding land markets, the issue is that on-farm investment in productivity-
increasing technology is seriously constrained by declining farm sizes, fragmented through inheritance. A
recent Government decree provides for minority title holders to cede their holdings to the majority (over 51
percent) but this has not yet been implemented”

European Union and bilateral aid. This may encourage the Bank to push money toward
Tunisia, a suspicion that is reinforced by the relatively large number of rural projects
overlapping closely in time and by theme. The time has come for the Bank to step back
and evaluate the long-term effectiveness of its program of assistance to the sector.

Box 2. Response from the Region (MNSRE) to OED’s Comment in Paragraph 7.4

 “The OED report suggests that the Bank needs to step back and evaluate the long-term effectiveness of its
program of assistance to the sector. On this last point, MNSRE has a slightly different view. True, we
agree that an evaluation of Bank lending for the rural sector is needed, but perhaps not for the same reason
as OED suggests. OED points to two main reasons : (i) the overlapping projects in time and by theme; (ii)
Tunisia’s access to cheaper and less conditional funding, which results in the Bank being less
discriminating than it should be in lending to Tunisia.

We do not perceive that our present portfolio, with four projects in the agriculture sector: Natural Resource
Management, Water Sector Investment, Agricultural Support Services and Northwest Mountainous Area
Development Project, represents a large number of projects, overlapping in objectives. The Water Sector
Investment addresses the crucial issue of agricultural water shortage in Tunisia and agricultural production
riskiness under difficult climatic circumstances. The Agricultural Support Services Project purports to
upgrade agricultural services with the perspective of trade liberalization. The Natural Resource
Management and Northwest Mountainous Area Development Project may appear as overlapping with their
objectives: improving NRM and reducing rural poverty in difficult marginal areas. However, the agro
ecological zones are different and above all, the type of implementing agency is different. Given the fact
that the CRDAs are present everywhere, the question has been whether in other marginal regions the
approach of ODESYPANO can be implemented by CRDAs, without having to create an Office?

On OED’s second point, the nature of the Bank's relationship with Tunisia is evolving. Precisely because
of the availability of other more attractive sources of funds and the Bank's comparative advantage in global
knowledge, there is a move beyond traditional lending. Tunisia looks to us to support innovative, often
pilot operations which are not so easily done with other donors (e.g., Natural Resource Management, Water
Sector Investment, Agricultural Support Services and Northwest III). Our Tunisian counterparts have
consistently highlighted that they want to continue borrowing from the Bank, for certain key operations,
even though they may have sources of easier and cheaper financial services, precisely because the Bank is
bringing qualified advice and support with its funding.

In our view, we need to think through our support to the agricultural sector and rural space in Tunisia on
the basis of the results of the competitiveness study. The study points towards a two-pronged evolution.

(i)      There will be considerable labor displacement from removing all farm subsidies and the opening
         of trade. In the long run, the displaced labor would be absorbed either by the rural economy
         through backward and forward linkages, or in the urban economy (secondary or tertiary sectors).
         However, during the transition period, measures of social protection will have to be provided.
         Indeed, the social and political burden from removing too rapidly farm protection may be even
         higher than mentioned in paragraph 1.6 of OED report. There are some 470 000 farms in Tunisia
         out of which 85 % earn less that $ 4500 a year, but employ between 2 and 7 persons per farm.
(ii)     The study also points out that there could still be a vibrant small-holder agricultural sector because
         of diversification and demand for quality agricultural products, in particular high value fruit and
         vegetables and processed products (wine, olive oil) for exports, for the tourist industry as well as
         for the domestic market (increasing demand for quality products due to increased incomes).

Given the above, the Bank team and the GOT need to direct their thinking towards a two-pronged approach
aimed at identifying:

(i)      the proper social safety net strategy to be implemented under cross-sectoral rural development

       programs that will address the welfare needs of the rural society and economy as a whole (such as
       under the Northwest Mountainous Areas Development Project)
(ii)   an investment strategy, to help small-scale farmers with potential to evolve towards more
       competitive production systems and produce quality agricultural products for which there is a
       market. Such a strategy would certainly include among other actions, (a) legislative and
       institutional reforms, decentralization and devolution of responsibilities to liberate private sector
       and civil society dynamism; (b) a rethinking of government services’ role to ensure the needed
       policy and regulatory framework; (c) land tenure reforms; as well as perhaps, (d) some well
       focused transition measures aiming at replacing public intervention with private activity so that
       privately financed solutions become viable. Such an approach has been started under the
       Agricultural Services Support Project and the Water Sector Investment Project but there is still a
       long way to go.

Lessons of General Relevance for Bank Operations

7.5    A piecemeal approach to reform—trying to effect small changes under the cover of
a number of investment operations—offers little leverage and is therefore unlikely to work.
In Tunisia, following the two agricultural structural adjustment operations, there has been
no lending vehicle that has allowed for sustained policy dialogue. This may be partly
because Tunisia has ample access to alternative, less conditional, sources of finance.

7.6      Policy reform has primarily been treated as an adjunct of investment, rather than
as a precondition for it: little attempt has been made to delay loan effectiveness until
agreed policy conditions have been met. Good illustrations of this are the Bank’s tacking
on of small land reform and rural finance initiatives to investment operations whose
primary objective lies elsewhere: land reform was tacked on to Northwest Mountainous
Areas Development (a successful pilot—but with no assured replicability) and the
Agricultural Sector Investment Loan (component dropped). The attempt to develop new
forms of rural financial intermediation—in the Northwest and National Rural Finance
projects—was also small-scale and ineffectual, lacking any government commitment. A
GTZ initiative in the Northwest—conducted outside the Bank-supported project—may
have been successful; but there is no government undertaking to scale it up: the message
is that the legal impediments to the mutualist experiment have to be tackled centrally—
and top-down—rather than through pilots.

7.7     Attempts to strengthen monitoring are futile if the performance indicators
monitored are not linked to the achievement of significant outcomes. The monitoring of
research and extension has improved; but most of the indicators are input based, rather than
outcome based. The push for annual increases in the indicative water tariff is only a
meaningful indicator of progress if there is a proportional increase in the tariff that users
actually pay. This is not the case: negotiation of the indicative tariff occupies a separate
track from negotiations made by each water user group. Monitoring the performance of
BNA’s portfolio does not enhance the capacity of BNA management to raise performance;
because key decisions on rescheduling are made by government, not by BNA.

7.8     Decentralization is an important objective; but it may be counterproductive if
resources are spread too thinly. Research and extension best exemplifies this: the number
of regional research poles envisaged probably exceeds the capacity to fill them; staff
shortages leave many of the local extension offices empty. ODESYPANO wisely chose
to target the quality of village development councils rather than the quantity—avoiding
the temptation to create as many councils as possible, which would have spread
supervision resources too thinly. The capacity of the Regional Agricultural
Commissariats to supervise water user groups—both in the irrigated perimeters and in the
rural drinking water supply schemes—shows signs of being overstretched.

7.9      Where there are multiple projects overlapping in time and by theme, it is
important to assess the overall impact of assistance: where there have been several
interventions which are cumulative in effect it is reasonable for evaluators to have higher
expectations of development effectiveness and to be correspondingly more demanding in
their search for evidence of effectiveness. With respect to the projects examined here, it is
difficult to attribute outcomes to specific operations. It is reasonable to expect that the
effect of these overlapping projects will be cumulative, meaning that toward the end of the
series, the outcomes expected should be more substantial than if the project had no close
forerunners. This implies that the evaluative bar should be raised over time. When
evaluating a series of overlapping projects, each with satisfactory outcomes, evaluators
must be prepared to rate the sequence as less than satisfactory if the expected cumulative
development impact has not materialized. This underlines the need to address the whole
program of assistance—the approach that will be followed by OED in the FY04 Country
Assistance Evaluation—rather than merely focusing on individual projects.
                                                      23                          Annex A

Annex A: Agriculture and Rural Development Overview of
There are three ecological zones. The North covers 25 percent of Tunisia’s land area and
is the most fertile part of the country. Rainfall is relatively high (400–1,000 mm/year).
The main products are milk, vegetables, grapes, citrus, meat, and cereals. The northwest
corner of this zone (an area of high poverty) contains productive natural pastures and
forests. The Tunisian dorsal (an extension of the Atlas mountains) separates the North
from the Center zone, which is bounded on the south by a series of large salt lakes
(shotts). The Center, which covers 15 percent of Tunisia, receives between 200 and 400
mm/year of rainfall, and the natural vegetation consists of esparto grass and large shrubs.
This zone is devoted mainly to cereals, olive trees, and pasture. The South accounts for
60 percent of the country’s land area. It is a pre-desert zone receiving less than 200 mm
of rainfall each year. The sparse vegetation consists mainly of small shrubs. There is
some extensive grazing, and some irrigated agriculture based on groundwater drawn from
wells.37 The population is becoming increasingly concentrated along the eastern littoral,
with 43 percent living in Tunis or along the mid-east and north-east coast.38

Over one-quarter of the land (5 million hectares) is cultivated, while just under one-
quarter (4.2 million hectares) is pasture or forest. One-third of arable land receives more
than 400 mm of rain per year (compared to 59 percent in Morocco). Only 7 percent of
cultivated land (345,000 hectares) is irrigated—a small proportion compared to the
Middle East average—and the output of rain-fed agriculture fluctuates sharply from year
to year according to rainfall. Around 20,000 hectares of agricultural land or pasture are
lost each year through erosion, desertification, salination, or urbanization.39

The agriculture sector has shrunk substantially in recent decades. Tunisia has a
population of almost 10 million, one-third of which lives in rural areas. The share of the
labor force in agriculture declined from 61 percent in 1961 to 25 percent in 1999.
Agriculture value added accounted for 12 percent of GDP in 2000, down from 16 percent
in 1990. Between 1990 and 2000, agriculture GDP grew at an average annual rate of 2.4
percent compared to economy-wide growth of 4.7 percent.40

Tunisia has undertaken reforms in agriculture since 1989. Subsidies for fertilizer, animal
feed, seed, irrigation, and mechanized services have been substantially reduced. The
supply of farm inputs, collection of produce, and the provision of mechanized plowing
and harvesting have been privatized. The role of private extension agents and
veterinarians has expanded in recent years. Progress has been slower in liberalizing food
marketing, the state remaining involved in cereals, olive oil, sugar, tea, coffee, tobacco,
and wine. Limited security of tenure, holding fragmentation, and land titling and

37. World Bank, Tunisia, Agricultural Sector Survey, September 29, 1982, p. 26.
38. Economist Intelligence Unit, Tunisia Country Profile 2002, p. 17.
39. Economist Intelligence Unit, Tunisia Country Profile 2002, p. 18.
40. World Bank, World Development Indicators, 2002.
                                                      24                                         Annex A

registration problems are issues that still await resolution. Water is a key constraint and
the importance of an integrated management of ground and surface sources has been
recognized.41 While all urban homes have piped water, the figure is only 24 percent in
rural areas. Some 5 percent of village homes are still more than 3 kilometers from their
nearest water source.42

Substantial progress has been made in reducing poverty—which is now primarily a rural
phenomenon.43 The percentage of persons under the national poverty line fell from 40
percent in 1960 to around 7 percent in the late 1990s. In 1995, the incidence of rural
poverty was 13.9 percent compared to 3.6 percent in urban areas, and more than 70
percent of the poor live in rural areas. The incidence of poverty is highest in the
mountainous north-west, and above the national average in the hilly and desert areas of
the center-west and the south. Poor rural households typically have access to land but
their holdings average only 2 hectares, are rarely irrigated and show low productivity.44

41. World Bank, Tunisia, Country Assistance Strategy, March 28, 2000, Annex B9a, p. 4.
42. Economist Intelligence Unit, Tunisia Country Profile 2002, p. 30.
43. M. Ayadi et al., “Putting robust statistical methods into practice: poverty analysis in Tunisia,” Swiss
Journal of Economics and Statistics, Vol. 137, No. 3, September 2001, pp. 463–482. The authors note that
“poverty in Tunisia is clearly a rural phenomena and this contradicts the finding of governmental
institutions,” p. 478.
44. World Bank, Tunisia, Country Assistance Strategy, March 28, 2000, p. 6.
                                                          25                                                 Annex B

Annex B: Results Matrices

Objectives. Improve the institutional and organizational framework needed to increase the effectiveness of research and

Component       Activities                Targets                        Outputs                    Outcomes
(a) Research
Improve         Set up IRESA to           IRESA to be created before     IRESA operational by       IRESA’s coordinating
programming     improve coordination      12/31/1990                     03/91                      powers limited, little
and             between research                                                                    leverage over central
budgeting       facilities                                                                          research institutes
Upgrade         Decentralize and          Strengthen central research    Central institutes         Regional research
Research        rationalize the           facilities                     better equipped            capacity remains
Facilities      administrative            Open 3 regional centers        Regional target            weak, poorly
                framework for             Close around 30 of the 46      changed: 7 PRRs to         administered
                research                  experimental stations.         be set up, of which 2
                                                                         in operation.
                                                                         No progress on
                                                                         station closures.
Strengthen      Recruit and train staff   Recruit 42 new staff, inc. 6   47 staff recruited, inc.   Staffing strengthened
staffing                                  agro-economists                3 agro-economists          but incentive to work
                                          75% to be posted outside       28% posted outside         outside Tunis limited
                                          Tunis                          Tunis
                                          243 man-months of training     57 man-months of
Strengthen      IRESA will                                               Seminars used to           Linkage still fairly
links between   disseminate research                                     disseminate research       loose; AVFA and
research and    results.                                                 results, with              IRESA do not
extension       A Research Council                                       participation of           collaborate in
                and an Extension                                         extension service;         preparing extension
                Council to be created                                    more ad hoc contact        messages
                in MinAg.                                                between two services.
(b) Extension
Equip central   Assign & train staff &                                   AVFA set up in 1991        Monitoring system in
monitoring      provide equipment                                                                   place, but remains
unit (AVFA)                                                                                         input oriented
Strengthen      Assign & train staff &    Support to 15 of the 23        Target met                 Benefits are
regional        provide equipment         CRDAs plus 104 sub-                                       dissipating owing to
centers                                   regional units (CTVs)                                     failure to replace staff,
(CRDAs)                                                                                             vehicles
Strengthen      Set up offices, provide   Upgrade 550 existing CRAs      60 CRAs improved           Many CRAs not
local units     vehicles and training     Build 54 new CRAs              40 new CRAs built          staffed, resources
(CRAs)                                                                                              spread too thinly
Rural           Introduce pilot           1993 target: 3 zones           Expanded to cover 8        Still covers 8 zones.
women’s         extension program                                        zones by 1996.             There are now 29
program                                                                  Demand-driven              female extension
                                                                         approach, high             agents, 12 in the field.
                                                                         Only 11 of 38 female
                                                                         extension workers
                                                                         work in the field.
Final Costs (US$ million, % of total)
Research 11.2m (54%)
Extension 9.7m (46%)
Total 20.9m (100%)*
*Investment costs only, no data on actual operating costs (estimated cost was US$12.9m).
                                                            26                                                 Annex B

Support growth through assuring availability and improving management of public resources […support public investment in
activities where the State has a justifiable role to play…improve the incentive framework…strengthen institutions, primarily the
Ministry of Agriculture.

Components         Activities                          SAR Targets          Outputs (ICR)                  Outcomes
(a) Investment
Irrigation and     Build hill dams.                                         15 dams completed.             Considerable expansion
water use          Bore tubewells.                     Bore 150             318 tube wells sunk.           of infrastructure with
efficiency         Boost perimeters.                   tubewells.           47 perimeters created,         large potential growth
                   Monitor groundwater.                                     covering 4,809 has.            impact.
                   Supply potable water.               Create 30            21 perimeters rehabilitated,
                                                       perimeters,          improving 2,750 ha.            Hill dams silt up rapidly
                                                       covering 2,500       57 meters installed.           and are not
                                                       ha                   44m cu.m. extra p.a.           economically viable.
                                                                            185 potable water projects,    Irrigation perimeters are
                                                                            231,000 benefited.             underused, reducing
Natural            Develop fisheries.                  /1                   36 ports upgraded              rate of return on
resource           Conserve soil & water.                                   416 flood control &            investment.
management         Consolidate farm holdings.                               groundwater recharge
                                                                            structures built,              Crop intensification
                                                                            34,000 ha of watershed         averages 123% on ASIL
                                                                            benefited                      perimeters (cf 90% on
                                                                                                           all perimeters).
Animal health      Give compulsory vaccinations.       Upgrade 19,162       2 animal health labs built.
and production     Boost disease surveillance.         has of               15,429 has upgraded, with      Improved rangeland
                   Improve public hygiene.             rangeland.           5,241 beneficiaries.           management, but
                   Upgrade rangelands.                                                                     heavily subsidy
                   Genetically enhance breeding.                                                           dependent.

(b) Policy Reform
Water sector      Raise CRDA water charges             /1                   Water tariff up by 9% pa,
                  Apply binomial tariff structure                           1994–98                        Indicative tariff raised;
Livestock sector Privatize veterinary services         c.110,000            54,617 contracts with          but tariff paid by water
                  Recover cost of services given       contracts with       private vets.                  users is much less, and
                  Set up disease surveillance          private vets for                                    often does not cover
                  network                              artificial                                          operations and
                  Strengthen Directorate for           insemination                                        maintenance costs.
                  Veterinary Public Hygiene
                  Strengthen National Program for                                                          Privatization of livestock
                  Genetic Improvement                                                                      services is proceeding
Fisheries sector Review incentive framework            /1                   EU certification obtained      slowly.
                  Introduce a Quality Assurance
                  Program (to boost access to EU                                                           Fish exports to EU
                  markets)                                                                                 stable but not
Soil and water    Submit Code for Water and Soil       /1                   100% of conservation           increasing.
conservation      Conservation to Chamber                                   works contracted.
                  Deputies                                                                                 Land reform component
Land reform       Strengthen legal framework for       /1                   AFA & ODESYPANO                was too small to have a
                  land consolidation                                        piloted attempts at            significant impact.   .45
                                                                            voluntary consolidation

(c) Institutional Development
Investment          Finance design and installation    /1
monitoring          of monitoring system

45. There is some disagreement about actual results. The completion report (p. 10) states that the project
“supported the successful development of a pilot intervention approach of voluntary consolidation of
fragmented parcels”; but Annex Table 2 on project costs shows that nothing was spent on the land reform
component (originally budgeted at US$600,000). The implementing agency (DGFE) was unaware of any
significant land reform initiative.
                                                             27                                                  Annex B

Decentralization    Study how to strengthen CRDAs       /1                   CRDA study dropped              CRDAs not
Investment          Various capacity building           /1                                                   strengthened
planning            measures.
User groups         Prepare strategy for promoting      /1                                                   Water user groups are
                    water user groups                                                                        gradually growing in
                                                                                                             strength, but CRDA
                                                                                                             supervision efforts are
Final Costs (US$ million, % of total)
Dams, 24.3m (14%)
Perimeters, 44.7m (25%)
Hydraulic studies, 41.4m (23%)
Drinking water, 36.4m (21%)
Livestock, 4.9m (3%)
Fisheries, 11.4m (6%)
Soil and water conservation, 12.4m (7%)
Land reform, 0.0m (0%)
Resettlement, 0.0m (0%)
Other, 1.5m (1%)
Total, 177.0 (100%)
/1 No targets were defined at appraisal. Items in italics cover activities where targets were set and outputs quantified.
CRDA Regional Commissariat for Agricultural Development
                                                            28                                                 Annex B

To alleviate poverty, improve the well-being of the region’s population, and to arrest degradation of the natural resource
base with active involvements of the population. To this end the project would: (i) promote measures to increase on-farm
productivity and off-farm income supporting activities; (ii) improve the management and productivity of range and farm
land; (iii) promote measures to reduce erosion, run-off, and reservoir sedimentation; (iv) improve social conditions of the
disadvantaged population by providing basic infrastructure and social services; and (v) promote increased involvement of
village organizations to take on development responsibilities.

Component        Activities                Targets                          Outputs                   Outcomes
(a) Agricultural Development
Livestock        Increase milk and                                          53,500 natural and        Improved techniques
production       meat production via                                        15,500 artificial
                 (a) artificial                                             inseminations.            Increased farm
                 insemination & (b)                                         Milk production from      incomes.
                 improved fodder                                            hybrid cows :1469
                 production                                                 kg/yr/cow.                Titling of remaining
                                                                            Meat production from      pilot area complete
                                                                            hybrid animals: 109
                                                                            kg carcass
Crop             Improved techniques:                                       Wheat 15.4 q/ha
production       wheat, barley, food                                        Barley 15.5 q/ha
                 legumes, olives
Dissemi-         Set up on-farm      1,500 on-farm crop                     609 plots
nation of        demonstration plots demonstration plots
results & new
Land con-      Surveys & mapping     None                                   2,100 ha reviewed,
solidation     (by Agence Fonciere                                          titles given for 1,200
pilot          Agricole)                                                    ha (900 ha pending)
(Increased     All of above          TD2,900/yr for small &                 TD1,633/yr for small
agricultural                         TD5,700/yr for medium                  & TD4,724/yr for
income)                              farmers                                medium
(b) Watershed and Rangeland Management
Conservation   8 measures to control Soil & water conservation              Soil & water              Interest in agro-
               erosion, conserve     works on 26,400 has;                   conservation works        forestry stimulated.
               water                 Increase perennial crops by            on 53,044 has;
                                     11,200 ha; 80% of works                Increase perennial        Works built by local
                                     executed by small local firms          crops by 19,718 ha;       firms, boosting
                                                                            100% of works             capacity
                                                                            executed by small
                                                                            local firms
(c) Income-Generating Activities
Provide        Set up Douar Credit                                          Component canceled
village micro- Union                                                        (DCU not legal)
Provide        Funding of small                                             657 applications          Rural finance remains
alternative    projects by (a) micro-                                       reviewed by BTS, 320      constrained
financing      credit from Banque                                           approved, 174
(after above   Tunisienne de                                                completed, 77
was            Solidarite (with NGO                                         underway;
canceled)      intermediation) and (b)                                      Fonds d’appui
               Fonds d’appui a l’API                                        supported 1
               (Approche                                                    beekeeping coop & 6
               participative integree)                                      soil & water
(d) Applied Research
               Support demand-                                              13 research themes        “Reasonable”
               driven short-term                                            contracted, 12            adoption rates (might
               research, contracted                                         completed. Highlights:    be higher if farming
               out to research                                              Improved implements       systems approach
               organizations                                                for animal-traction,      were used)
                                                              29                                                 Annex B

                                                                              improved pasture
                                                                              seed, technique for
                                                                              measuring impact of
                                                                              soil and water
                                                                              conservation, guide
                                                                              for preparing
                                                                              Development Plans
(e) Rural Infrastructure
Rural roads     Build new roads and          300 km built; 650 km             383 km built; 638 km      Responsibility for
                rehabilitate existing        rehabilitated                    rehabilitated             maintenance remains
                ones in isolated rural                                                                  undefined.
Water supply    Various measures             Identify and equip 35 springs;   50 springs identified     7,800 people given
                                             Sink 28 wells; Build 19 AEP      and equipped; 7 wells     better water supply,
                                             potable water systems; Build     sunk (shortfall           5,400 from AEP
                                             15 hill dams                     compensated by            systems, 2,400 from
                                                                              building of 477 water     tanks.
                                                                              tanks, exceeding mid-
                                                                              term estimate of 300);
                                                                              4 hill dams built
Schools &         Build schools and          Reduce illiteracy rate to 45%    Targets cut because       Illiteracy rate cut from
health            health centers                                              of high unit costs of     60 to 44%.
centers                                                                       construction

                                                                              Achieved: 50% of
                                                                              MTR target for N of
                                                                              schools built; 72% of
                                                                              MTR target for N of
                                                                              health centers
(f) Institutional Strengthening
Training          Training (a) for village                                    980 beneficiaries         Steady progress in
                  development                                                 trained, including 160    improving capacity of
                  committees and (b)                                          young women; 650          village development
                  ODESYPANO staff                                             staff trained             committees and
M&E               Develop management                                          2 surveys (1998 and       community planning
                  tools, inc. GIS, MIS                                        2000) to measure          process.
                  and audit systems                                           progress in relation to
                                                                              1996 baseline.
Studies            Various, including use                                     9 studies completed
                   of participatory
                   approach, land
Final Cost (US$ million, % of total)
Rural infrastructure, 21.8m (46%)
Watershed/rangeland,18.3m (39%)
Agricultural development, 3.7m (8%)
Institutional strengthening, 3.0m (6%)
Applied research, 0.3m (3%)
Income generation, 0.0m (0%)
Total, 47.1m (100%)
                                                            30                                                 Annex B


To assist Tunisia in promoting creditworthy private investment in rural areas and in strengthening, on a sustainable basis, the
financial viability and institutional reform process of the Banque Nationale Agricole.

Components         Activities                          SAR Targets          Outputs (ICR)                  Outcomes
(a) Credit Line
                   Finance farm and off-farm           Inc. US$15.0m        No data on counterpart         Weak loan recovery
                   investments, including agro-        to large farmers,    funding, but of Bank loan,     (39% average) suggests
                   industry, rural coops, fisheries,   US$10.5m to          US$11.7m went to large         many non-viable
                   women’s enterprise, housing         small farmers        farmers, US$12.5m to           proposals were financed
                                                       and US$15m to        small farmers and
                                                       agro-industry        US$18.7m to agro-industry
(b) BNA Institutional Development
                     Improve management, with           Various portfolio performance indicators were  No sign of sustained
                     decentralization to branch         developed at mid-term and monitored            improvement in portfolio
                     offices; develop plan to absorb    subsequently (see Annex C, Table C10)          quality, reflecting BNA’s
                     BNA arrears; separate                                                             lack of autonomy in
                     government credit lines from                                                      relation to government
                     regular operations
(c) Study
                     Assess the feasibility of village-                      Original study dropped,   Challenge of developing
                     based group-lending, including                          substituted by a study on alternatives to BNA
                     existence of informal financial                         financing of small-scale  finance not met
                     markets and legal and                                   agriculture
                     regulatory framework
Final Costs (US$ million, % of total)
No data on final costs. Of money disbursed from Bank loan, US$47.2m financed the credit line, and US$0.1m was devoted to
institutional development.
                                                          31                                               Annex C

Annex C: Supplemental Tables

Table C1. Key Project Ratings Comparing Tunisia with Elsewhere
FY1991–03             N of projects     Outcome            Sustainability      Institutional        Borrower
                                                                               Development          Performance
                                                                               Impact               (Implementation)
                                        % Satisfactory     % Likely            % Substantial or     % Satisfactory
Tunisia: Rural               14               86%                64%                  50%                   71%
Tunisia: All                 41               83%                73%                  43%                   70%
MNA: All                     242              68%                53%                  32%                   63%
Bank-wide: All         2,835                  70%                53%                  37%                   64%
Source: OED Ratings database.

Table C2. Ratings of Projects Relevant to this Evaluation
                              Rating    Outcome*           Sustainability       Institutional       Borrower
                              Source                                            Development         Performance
                              (Date)                                            Impact              (Implementation)
Third Agricultural Credit       PAR       Satisfactory         Uncertain           Substantial          Satisfactory
(L1885)                        Jun 93
Northwest Rural                 PAR       Marginally           (Unlikely)           (Modest)          (Unsatisfactory)
Development (L1997)            Jun 93    Unsatisfactory
Northwest Agricultural          PCR      Unsatisfactory         Unlikely           Negligible             Highly
Production (L2502)             Mar 94                                                                  Unsatisfactory
Agricultural Sector             PAR       Satisfactory          (Likely)          (Substantial)        (Satisfactory)
Adjustment (L2754)             Feb 93
Fourth Agricultural Credit      PAR       Satisfactory         Uncertain           Substantial          Satisfactory
(L2865)                        Jun 93
Second Agricultural             ICR       Satisfactory           Likely             Modest              Satisfactory
Sector Adjustment              Feb 96
Agricultural Research          ICR        Moderately           Uncertain            Modest            Unsatisfactory
& Extension (L3217)           Dec 97      Satisfactory
Agriculture Sector              ICR       Satisfactory       Highly Likely        Substantial          Satisfactory
Investment (L3661)             Jun 01
Northwest                      ICR        Moderately             Likely           Substantial          Satisfactory
Mountainous Areas             Dec 01      Satisfactory
Development (L3691)
National Rural Finance         ICR      Unsatisfactory          Unlikely            Modest            Unsatisfactory
(L3892)                       Dec 01
Second Agricuture               PSR       Satisfactory            N/A                 N/A                   N/A
Sector Investment              Oct 02
Agriculture Services           PSR        Satisfactory            N/A                 N/A                   N/A
(L7063)                       Sep 02
Northwest Mountainous           PSR          Satisfactory          N/A                 N/A                   N/A
and Forestry Areas             Oct 02
Development (L7151)
Source: OED Ratings Database. PAR Performance Audit Report; PCR Project Completion Note; ICR Implementation
Completion Report; PSR Project Status Report; N/A Not Applicable.
* In case of PSR, outcome refers to progress toward the project’s development objectives. Ratings in parentheses are
imputed: they are not explicit in the evaluation report.
                                                32                                     Annex C

Table C3. Timeline of Tunisia Rural Projects, 1980–2003
                                               BANK FINANCIAL YEAR (1980–2003)
                 8 8   8   8   8   8   8   8   8  8    9  9   9   9    9   9   9   9   9   9   0   0   0   0
                 0 1   2   3   4   5   6   7   8  9    0  1   2   3    4   5   6   7   8   9   0   1   2   3
Third Ag.
Credit (L1885)
Ag. Credit
Research &
National Rural
and Forestry
                                                        33                                              Annex C

Table C4. Research and Extension Indicators
                                       1997          1998          1999          2000            2001       2002
N of new technologies/N of              NA            NA            NA           21%             25%        16%
researchers (%)
Mobility of extension agents (Days      1.5           1.6           1.9           1.9            1.9         1.9
per agent per week with access to a
Time devoted to extension/Time at      55%           55%           53%           50%             50%        50%
work (%)
Source: Aide-Memoire, ICR mission, ASIL2 project, March 2003.

Table C5. Northwest Mountainous Areas Development Project: Cost by Component
Component                                           (1) Estimated Cost         (2) Actual Cost           (2)/(1)
                                                        US$ million              US$ million              (%)
Agricultural Development                                    6.0                      3.7                  62%
Watershed & Rangeland Mgmnt.                                12.3                    18.3                 149%
Income Generating Activities                                1.5                      0.0                   0%
Applied Research                                             1.3                     0.3                  23%
Rural Infrastructure                                        21.3                    21.8                 102%
Institutional Strengthening                                 5.3                      3.0                  57%
Unallocated                                                  3.0                     0.0                   0%
Total                                                       50.7                    47.1                  93%
Source: Implementation Completion Report,

Table C6. Capacity of ODESYPANO Development Committees
Development Status          Committees              Community             Area Covered           Population Covered
of Committee                                       Development             (Hectares)
                                                 Plans Completed
Good                     114 (44%)                   89 (74%)             148,319 (46%)            103,290 (54%)
Moderate                 52 (20%)                    32 (26%)              72,106 (22%)             39,811 (21%)
Poor                     94 (36%)                      0 (-)              104,316 (32%)             46,959 (25%)
Total                   260 (100%)                  121 (100%)            324,741 (100%)           190,060 (100%)
Source: ODESYPANO, March 2003
                                                              34                                              Annex C

Table C7. Performance of Selected Irrigation Perimeters
Perimeters/          Gross           Revenue          Tariff          % of irrigated    Intensification
Farm models          margin          per m3           (dinar/m3)      area in
                     (dinar/ha)      (dinar)                          cereals and
Lezdine 1            3,852           0.896            0.130           57%               100%
                     1,814           0.703            0.130           92%               100%
Ferinana 6           427             0.160            0.062           NA                115%
                     599             0.244            0.062           NA                NA
                     503             0.263            0.062           NA                NANA
Eddissa              428             0.178            0.020           67%               84%
Negagta              564             0.195            0.075           0%                150%
                     2,195           0.519            0.075           12%               141%
                     519             0.212            0.075           0%                144%
El Aguila            1,592           0.305            0.018           57%               117%
                     951             0.208            0.018           47%               158%
                     310             0.070            0.018           30%               147%
Teboulba             9,910           4.731            0.150           0%                120%
                     7,377           2.598            0.150           0%                127%
                     6,970           2.753            0.150           0%                131%
Beni Othman          560             0.246            0.100           0%                164%
                     882             0.455            0.100           0%                129%
Sidi Amor B S        724             0.355            0.050           NA                NA
                     527             0.216            0.050           NA                NA
                     695             0.295            0.050           NA                NA
Sidi Aissa           1,917           0.458            0.075           0%                133%
                     2,032           0.508            0.075           13%               123%
                     1,753           0.616            0.075           0%                100%
Bougossa II          2,771           0.857            0.070           100%              100%
                     1,431           0.532            0.070           80%               125%
                     1,177           0.475            0.070           70%               122%
El Azima             1,016           0.770            0.120           74%               100%
                     745             0.660            0.120           72%               100%
Mnasria              2,122           0.617            0.040           NA                NA
                     1,210           0.482            0.040           NA                NA
                     1,908           0.571            0.040           NA                NA
El Hajeb             955             0.362            0.045           NA                NA

N of Farm            32             32             13             22                    23
Models                                             (Perimeters)
Mean                 1,889          0.703          0.070          35%                   123%
Variance*            117%           134%           59%            104%                  17%
Source: Etude d’evaluation d’impact socio-economique des perimeters irrigues, 2003
NA Not available
* Variance is standard deviation/mean

Table C8. Financial Rate of Return, Bourgossa II Perimeter, Jendouba.
                                                         Farm Model 1            Farm Model 2             Farm Model 3

Cereals (ha)                                                     -                        2.0                   3.5
Forage (ha)                                                     0.5                       2.0                   2.5
Olives (ha)                                                      -                         -                    0.3
Fruit & vegetables (ha)                                          -                        1.0                   2.2
Incremental gross margin per hectare, US$/1                     804                      958                   1,940
Investment cost (tube well & works), Year 1, US$,              396,720                396,720                 396,720
Maintenance cost, Years 2–20, US$                              3,316                    3,316                  3,316
FRR at full use (80 hectares)                                  13%                       16%                   28%
FRR at current use (30 hectares)                                1%                        3%                   12%
Source: Jendouba Monitoring Report, 2002; Impact study, 2003
/1 Based on the difference before and after the project, with project benefits constant from year 4 to year 20
/2 Assuming present level of intensification (each hectare cropped 1.25 times per year).
                                                           35                                                Annex C

Table C9. Rainfall and BNA Loan Recovery, 1994–2001
Location*            Annual Rainfall (mm), 1994–2001       Annual Loan Recovery (Dinars), 1994–2001        Correlation
                       Mean              Variance**                Mean                Variance**         Coefficient***
Tabarka                 952                 0.239                  8,591                 1.167                 0.39
Bizerte                 562                 0.247                 15,393                 0.641                 0.42
Jendouba                482                 0.237                 43,204                 0.454                 0.80
Thala                   364                 0.234                 13,086                 0.967                -0.16
Monastir                318                 0.491                 11,558                 1.035                 0.18
Kairouan                277                 0.422                114,096                 0.420                -0.05
Sidi Bouzid             227                 0.264                 10,506                 0.762                 0.74
Sfax                    211                 0.422                 65,899                 1.057                 0.28
Gabes                   163                 0.448                 11,150                 0.665                -0.02
Gafsa                   140                 0.514                 31,963                 0.401                -0.01
Medenine                134                 0.343                  4,583                 0.907                 0.26
Tozeur                   75                 0.453                874,407                 1.002                 0.35
Source: Rainfall from Annuaire des Statistique Agricoles, 2001; loan recovery data from BNA
* Meteorological station and BNA branch office. **Variance=Standard deviation/Mean. ***Rainfall correlated with recovery
of agricultural for each of the seven years, 1994/95 to 2000/01.

Table C10. BNA’s Loan Portfolio Quality Indicators
                                   12/31/1999         12/31/2000               12/31/2001               09/30/2002
                                      Base         Target     Actual        Target     Actual        Target     Actual
<360 days*                          5.9%           <5%           4.6%       <5%           4.9%       <5%           6.1%
>360 days*                         24.4%          <22.4%        20.0%      <20.4%        19.1%      <18.4%        20.2%
Portfolio at Risk
Overall                             39.5%         <36.5%        34.5%       <33.9        30.0%       <30.9         NA
Commercial & Industrial Loans      27.1%          <27.1%        19.4%       <2000        15.6%       <2001         NA
Agricultural Loans                 66.9%          <66.9%        68.0%       <2000        67.0%       <2001         Na
Source: BNA
*Non-budgetary loans. NA Not available.

Table C11. Bank Preparation and Supervision Costs
Project                                              Preparation Costs                      Annual Supervision Cost
                                                   Project (Region Mean)*                   Project (Region Mean)*
                                                           US$’000                                  US$’000
Ag. Research & Extension                                212.1 (NA)**                              65.1 (43.0)
Northwest Development                                   318.2 (379.0)                             91.7 (49.4)
Ag Sector Investment                                    457.6 (379.0)                             96.9 (49.4)
National Rural Finance                                  539.0 (513.8)                             74.6 (53.5)
*Mean cost for rural projects in MNA region (preparation costs refer to FY of project approval, supervision costs refer to
average annual cost, for relevant years in period FY93-FY99).
** No data for rural project preparation costs in FY90, when this project was approved; in FY91 the mean preparation cost
for agriculture projects in MNA was US$243, 800).
Source: ICRs; Region means taken from OED, Rural Development: From Vision to Action? (Phase II), June 22, 2000,
Annex E.
                                                     36                                      Annex D

  Annex D: Basic Data Sheets


  Key Project Data (US$ millions)
                                 Appraisal                      Actual or          Actual as % of
                                 estimate                    current estimate    appraisal estimate
IDA Credit                        25.9                             15.3                 59
Cofinancing                         -                               1.0                  0
Government                         8.1                              4.6                 57
Total Project Costs               34.0                             20.9                 61

  Project Timetable
                                                     Original                         Actual
    Identification/Preparation                   August 23, 1988                 August 23, 1988
    Appraisal                                   September 23, 1988              September 23, 1988
    Approval                                    September 24, 1990              September 24, 1990
    Effectiveness                                 May 21, 1991                    May 21, 1991
    Mid-term review
    Credit closing                                June 30, 1996                   June 30, 1997

  Staff Inputs (staff weeks)
                                                   Actual weeks                   Actual US$000
    Through appraisal                    25.9                                          66.7
    Negotiations to Board                60.0                                         145.4
    Supervision                                     118.7                             390.6
    Completion                                         4.0                             15.7
    Total                                           208.6                             618.3
                                                    37                                             Annex D

Mission Data
                                                                                            Performance rating
                          Date           No. of
                                                     Specializations represented       Implementation Development
                       (month/year)     persons
                                                                                           status      objectives
 Identification/      July 7, 1987 -
 Preparation/        August 23, 1988
 Supervision 1       April -May 1991      1                      A                          S               S
 Supervision 2        October 1991        2                    A, R                         HS              S
 Supervision 3          July 1992         3                   A, R, F                       HS              S
 Supervision 4       March-April 1993     3                   A, R, F                        S              S
 Supervision 5       September 1993       3                   A, R, F                       n.a            n.a
 Supervision 6         April 1994         4                A, R, F, M&E                     HS              S
 Supervision 7        October 1994        2                    A, Ec                         S              S
 Supervision 8
                        June 1995         2                    A, Ec                        S               S
 (Mid-term review)
 Supervision 9         October 1996       3                   A, Ec, F                      S               S
 ICR                    May 1997          2                    A, Ec                         -              -

Specializations represented: A: Agronomics; Ec: Economics; F: Financial Analyst; R: Research; M&E:

Performance ratings: HS: Highly satisfactory; S: Satisfactory; n.a.: Not Applicable.
                                               38                                          Annex D


  Key Project Data (US$ millions)
                                 Appraisal                   Actual or          Actual as % of
                                 estimate                 current estimate    appraisal estimate
IDA Credit                        120.0                        115.8                  97
Cofinancing                          -                           -                     -
Government                         81.5                        61.3                   75
Total Project Costs               201.5                        177.1                  88

  Project Timetable
                                               Original                            Actual
    Identification/Preparation                                                October 23, 1992
    Appraisal                                                                  May 28, 1993
    Approval                                                                 November 18, 1993
    Effectiveness                                                              April 13, 1994
    Mid-term review                                                           October 1, 1996
    Credit closing                           June 30, 1999                   December 31, 2000

  Staff Inputs (staff weeks)
                                             Actual weeks                      Actual US$000
    Identification/Preparation                   40.7                              183.1
    Appraisal/Negotiation                        61.0                              274.5
    Supervision                                 145.4                              654.1
    ICR                                          17.3                               77.9
    Total                                       264.4                             1189.6
                                                    39                                        Annex D

Mission Data
                                                                                     Performance rating
                         Date           No. of
                                                  Specializations represented   Implementation Development
                      (month/year)     persons
                                                                                    status      objectives
 Identification/     November 1992        4            TM, IR, IS, AMS

 Appraisal/             June 1993         6       TM, LFS, IRS, LRS, HS, IS
 Supervision 1         April 1994         2                      Es                  HS                S
 Supervision 2       November 1994        2                 AE, LFS                   S                S
 Supervision 3        October 1995        4          TM, LFS, SCS, LTS                S                S
 Supervision 4          May 1996          4              TM, LFS, IRS                 S                S
 Supervision 5       November 1996        2                 TM, LFS                   S                S
 Supervision 6         April 1997         4            TM, IR, ES, DRS                S                S
 Supervision 7          May 1998          4            TM, LFS, RI, DS                S                S
 Supervision 8        October 1998        2                TM, GWS                    S                S
 Supervision 9          May 1999          4               TM, AS, IRS                 S                S
 Supervision 10      December 1999        4               TM, E, S, IR                S                S
 Supervision 11         July 2000         4              TM, IR, PMS, A               S                S
 Supervision 12      December 2000        4              TM, S, IR, IRS               S                S
 ICR                 December 2000        4              TM, S, IR, IRS

Specializations represented: TM=Task Manager; IR=Irrigation Speciation; IS=Institutional Specialist;
AMS=Agribusiness and Marketing Specialist; LFS=Livestock & Fisheries Specialist; IRS=Irrigation
Research Specialist; LRS=Land Reform Specialist; HS=Hydraulics Specialist; E=Economist;
AE=Agriculture Economist; SCS=Soil Conservation Specialist; LTS=Land Tenure Specialist;
ES=Environmental Specialist; DRS=Drainage Research Specialist; RI=Resettlement Irrigation;
DS=Drainage Specialist; GWS=Ground Water Specialist; AS=Agriculture Specialist; S=Sociologist;
PMS=Project Management Specialist; A=Agronomist.

Performance ratings: HS: Highly satisfactory; S: Satisfactory.
                                               40                                      Annex D

  (LOAN 3691)

  Key Project Data (US$ millions)
                                 Appraisal                Actual or         Actual as % of
                                 estimate              current estimate   appraisal estimate
IDA Credit                        27.5                       24.7                302
Cofinancing                        3.7                        4.6               124.3
Government                        19.5                       17.8               246.2
Total Project Costs               50.7                       47.1               672.5

  Project Timetable
                                               Original                        Actual
    Identification/Preparation                                              June 19, 1992
    Appraisal/Negotiation                                                   June 9, 1993
    Approval                                                              December 23, 1993
    Effectiveness                            April 30, 1994                August 23, 1994
    Mid-term review                          June 30, 1996                   July 5, 1997
    Credit closing                           June 30, 2000                  June 30, 2001

  Staff Inputs (staff weeks)
                                             Actual Weeks                  Actual US$000
    Identification/Preparation                 106.4                           118.7
    Appraisal/Negotiation                       70.0                           199.9
    Supervision                                178.0                           618.9
    ICR                                         10.0                            30.0
    Total                                      364.4                           967.1
                                                  41                                         Annex D

Mission Data
                                                                                     Performance rating
                      Date          No. of       Specialization represented
                   (month/year)    persons                                     Implementation   Development
                                                                                   status        objectives

Identification/     June 1990        7          A, E, RI, EX, EE, AL, W
Appraisal/          June 1993        7             E, E, AL, EX, U(3)
Supervision 1       May 1994         4                 NRM, A, E, S                  S                 S
Supervision 2     December 1994      2                     E, S                      HS                HS
Supervision 3       April 1995       3                  NRM, A, E                    S                 S
Supervision 4       May 1996         2                    OA, E                      S                 S
Supervision 5      January 1997      3                 E, AL, NRM                    S                 S
Supervision 6     June/July 1997     4                 OA, E, A, FN                  S                 S
Supervision 7       May 1998         3                 OA, NRM, A                    S                 S
Supervision 8     February 1999      5            NRM, OP, AL, E, G                  S                 S
Supervision 9     September 1999     3                  AE, OP, E                    S                 S
Supervision 10      May 2000         3                  NRM, E(2)                    S                 S
Supervision 11     January 2001      4                 NRM, E, S, W                  S                 S
ICR                March 2001        4                 E, AL, FN, S                  S                 S

Specializations represented: AE=Agricultural Economist; E=Economist; EX=Extension Specialist; RI=
Rural Infrastructure Specialist; F=Forestry Specialist; FN=Finance Specialist; NRM=Natural Resources
Management Specialist; OP=Operations Officer; P-Procurement Specialist; U=Unidentifiable;
AL=Agronomist/Livestock Specialist; S=Sociologist; OA=Operation Analyst; G=GIS Specialist;
W=Women Micro-enterprise Specialist; MTR=Mid-term Review; N.A= Not Applicable.
                                                  42                                    Annex D


  Key Project Data (US$ millions)
                                 Appraisal                  Actual or          Actual as % of
                                 estimate                current estimate    appraisal estimate
IBRD Credit                       65.0             No data in ICR
Cofinancing                      319.5
Government                        35.5
Total Project Costs              420.0

  Project Timetable
                                                  Original                        Actual
    Identification/Preparation                                              November 18, 1994
    Appraisal/Negotiation                                                   November 30, 1994
    Approval                                                                  May 23, 1995
    Effectiveness                             October 4, 1995               September 21, 1995
    Mid-term review                            April 30, 1997                October 20, 1998
    Credit closing                           September 30, 1999               June 30, 2001

  Staff Inputs (staff weeks)
                                               Actual Weeks                   Actual US$000
    Identification/Preparation                     N.A.                            N.A.
    Appraisal/Negotiation                          N.A                            539
    Supervision                                    N.A                            373
    ICR                                            N.A                             18
    Total                                                                         918
                                                    43                                          Annex D

Mission Data
                                                                                        Performance rating
                       Date         No. of        Specialization represented
                    (month/year)   persons                                        Implementation   Development
                                                                                      status        objectives

Identification/      July 1991        4             BS, ACS, CIS, RFS
Appraisal/          March 1995        5
Supervision 1     November 1995       4              FA, MS, ACS, BS                    S               S
Supervision 2        May 1996         4              FA, MS, ACS, BS                    S               S
Supervision 3        July 1997        2                    FA, FSS                      S               S
Supervision 4       March 1998        3                   FA, FSS(2)                    S               S
Supervision 5      October 1998       3                  FA, FSS, PSD                   S               S
Supervision 6        June 1999        2                   PSD, FSS                      U               U
Supervision 7     September 1999      1                      FSS                        S               S
Supervision 8        June 2000        1                      FSS                        S               S
Supervision 9     November 2000       2                   FSS, RDE                      S               S
ICR                     **

** Note: ICR mission planned for 9/01 was cancelled due to the terrorist attacks of September 11, 2001, in
the United States.

Specializations represented: BS=Banking Specialist; ACS=Agricultural Credit Specialist; CIS=Crop
Insurance Specialist; RFS=Rural Finance Specialist; FA=Financial Analyst; MS=Micro enterprise
Specialist; FSS=Financial Services Specialist; PSD=Private Sector Development Specialist; RDE= Rural
Development Expert.
Performance ratings: S: Satisfactory; U: Unsatisfactory.

To top