Document of The World Bank

FOR OFFICIAL USE ONLY Public Disclosure Authorized

Report No. 6245 Public Disclosure Authorized

PROJECT PERFORMANCE AUDIT REPORT

BOTSWANA

SECOND LIVESTOCK DEVELOPMENT PROJECT

(LOAN 1497-BT) Public Disclosure Authorized

June 13, 1986 Public Disclosure Authorized

Operations Evaluation Department

This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. ABBREVIATIONS

AMA - Agricultural K.,Lagement Associations APRU - Animal Production Research Unit BLDC - Livestock Development Corporation BMC - Botswana Meat Corporation CGC - Communal Grazing Cell CGU - Communal Grazing Unit DAH - Department of Animal Health DWA - Department of Water Affairs EDF - European Development Fund ERR - Economic Rate of Return FA0 - Food and Agriculture Organization FMD - Foot and Mouth Disease GOB - Government of Botswana ILCA - International Livestock Center for Africa LP1 - (First) Livestock Development Proje.:t (Credit 325-BT) LP2 - Second Livestock Development Project (Loan 1497-BT) LPCU - Livestock Project Coordinating Unit - LPMU - Livestock Project Management Unit LUPAGS - Land Use Planning Advisory Groups MOA - Ministry of Agriculture M&E - Monitoring and Evaluation MFDP - Ministry of Finance and Development Planning NDB - National Development Bank NLMLP - National Land Management and Livestock Project NTRP - National Trek Route Policy OED - Operations Evaluation Department PC - Project Coordinator PCR - Project Completion Report PMU - Project Management Unit PPAM - Project Performance Audit Report PSC - Project Steering Committee REU - Ranch Extension Unit RMTC - Ranch Management Training Center RSU - Rural Sociology Unit SIDA - Swedish International Development Association TGLP - Tribal Grazing Land Policy

CURRENCY EQUIVALENTS

Name of Currency (Abbreviation) Pula (P)

Currency Exchange Rate: Appraisal Year Average US$1.00 = P 0.83 Intervening Years Average US$1.00 = P 0.97 Completion Year Average US$1.00 = P 1.28

FISCAL YEAR BORROWER

January 1 - December 31 ONLY THE WORLD BANK FOR OFFICIAL USE Washington, D.C. 20433 U.S.A.

Office f Oirector-Cwal Operations aluation

June 13, 1986

MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT

Subject: Project Performance Audit Report Botswana Second Livestock Development Project (Loan 1497-BT)

Attached, for information is a copy of a report entitled "Project Performance Audit Report on Botswana Second Livestock Development Project (Loan 1497-BT)" prepared by the Operations Evaluation Department.

Attachment

This document has a restricted distribution and may be used by recipients only in the performance of their offcial duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY

PROJECT PERFORMANCE AUDIT REPORT

BOTSWANA

SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

TABLE OF CONTENTS

Page No.

Preface *** ******* ******* ********* ******** * Basic Data Sheet ...... *************...... 11 Evaluation Summary *...... o.* iv

PROJECT PERFORMANCE AUDIT MEMORANDUM

I. PROJECT SUMMARY ...... 1I......

Project Description 1I...... Project Implementation ...... 3 Projectblts ...... 6

II. ASSESSMENT OF PROJECT IMPACT AND RATIOALE ...... 8

ANNEXES:

1. Geographical Distribution of Project Sites ...... 13 2. Project Achievements **************************************** 14 3. Follow-up Project ************************************ 16

Regional Transmittal Memo ..... *****************************... 21

PROJECT COMPLETION REPORT

I Background ...... 27 II. Project Formulation *********** * *e****** 27 111. Project Implementation *************************** 33 IV. The Impact of Management on Project Achievement ...... 41 V. Agricultural Impact *********************************** 44 VI. Tribal Grazing Land Policy ...... 48 VII. Changes in Repeater Projects ...... 49 VIII. Bank Performance ****************************************. 51 X1. Conclusion *********************************************** 52

Mas:

IBRD No. 12825R Second Livestock Development Project IBRD No. 12824 Soil, and Transportation Network IBRD No. 18766 National Land Management and Livestock Project

This drcument has a restricted distribution and may be used by recipients only in the performance of th. i official duties. its contents may not otherwise be disclosed without World Bank authorization. - i -

PROJECT PERFORMANCE AUDIT REPORT

BOTSWANA

SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

PREFACE

This is a project performance audit on the Botswana Second Livestock Project for which Loan 1497-BT was approved in December 1977 in the amount of US$6.5 million and was declared effective November 1978. The Loan closed June 30, 1984 after one year's postponement with an undisbursed balance of US$.48 million which was cancelled in March 1985. The project was also cofinanced by the European Development Fund (EDF) for US$0.3 million and the International Livestock Center for Africa (ILCA) for US$1.26 million.

The audit report consists of a Project Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) prepared in July/August 1985 by the Ministry of Agriculture (MOA), Government of Botswana. The PCR was submitted to OED by the Region on February 25, 1986. The Region is in broad agreement with the PCR but has made two comments, one regarding lessons learned and one regarding benefits. The Region's transmittal memorandum has therefore been appended to the PPAK and immediately precedes the PCR.

The PPAM is based on the Appraisal Report (No. 1630-BT dated October 31, 1977), the President's Report (No. P-2140a-BT dated October 31, 1977) and the Loan Agreement of December 30, 1977. Bank files on the project, including corresponding with the Borrower, were fully consulted and Bank staff associated with the project were interviewed. In addition, an audit mission visited Botswana in November 1985 to discuss the project with officials of MOA (which was the executing agency) and with staffs of the Project Management Unit (PMU), the Animal Production Research Unit (APRU), the Ranch Extension Unit (REU), the Rural Sociology Unit (RSU), and the National Development Bank (NDB), all of which played important roles in implementing the project's principal components.

The audit finds the PCR detailed and accurate in its description of project achievements and shortcomings, and the audit generally supports its principal findings. The audit concludes that quantifiable benefits under the project mnst be considered highly conjectural at this time and they would be, in any event, extremely low, resulting in an ERR Thich would in all probability be inconsequential. On the other hand, the audit is convinced that the physical, institutional, training and technical assistance achieve- ments under the project were sufficiently significant in their own right to justify the project, particularly taking account of the novel and innovative nature of the complex local land reform efforts which the project was - ii - designed to support. The audit highlights the many difficulties which have surfaced in the last ten years during which Government has promoted its land reform policies, particularly those at the farm level, and concludes that the future will be painstakingly difficult with a slow direct benefit buildup and low anticipated ERR in the early stages.

The draft report was sent to the Borrower for comments on March 21, 1986. However, none has been received.

The audit mirsion wishes to thank the staff of the Government of Botswana, MOA, PMU and NDB for the valuable assistance and support they pro- vided during its visit to Botswana. - l1 -

PR0t80 RA~ AUD2T#P8RT 80T85ANA

8800 LIVEST?CKDIV8LOI'MENT 110JECT (i0AN 1497-8T) sAdIC DATA8m1u?

_~Y9AAPAJTCT Actual or Actual as 8 Appralal etiaated of Appraisal Istisate Atual ___i__ate

Total Projact Cont (US8 aillion) 1.4 8.1 60 ton aount (08 willto) 6.5 6.02 93 Cotiancing EDP 0.3 ILCA 1.26 Date ~tyica o nponets CoMpleted 06/83 06/84 120 Proportion Th~n Complated <1) 100 80 Eteolc Rate of Return (8> 13 0 Finaneial Perfornance fair Marginal inosttutional 9erfor~ance Satiafattory Satisfactory 8TAOKINPU18 (staff eek)

Yfis 7176 'n7 Me78 Mr79 VM8 Fu8l 118 1781 118 MS Total Ideatiftcation/Preparation 5.7 19.1 3.4 4.0 34.2 Appratsal 45.1 7.0 52.1 Negotiation 7.8 7.8 sopervicon 7 0 101å 7.5 12 9 16 1 1.6 6 4.0 76.1 Total T71T Ift, flå rt ifv ffIr TE12 1 f w

CI~IULATIVEDIBUR~81 Ts

Appråtsal ~tiaate (U88 gillion) 0.04 1.25 2.23 4.00 6.00 6.50 - Actual (Us$ million) 0 0.03 0.37 0.90 1.60 3.28 6.02 Actunt a %of Ratimte 0 2.4 16 23 27 81 93 Date of Ftnal DIGbursemat March 7, 198$

POJ&CTS DATM1

Original Estinate Revisons AtUa Firat Neatioed In fil* or Timetable 01175 Government equest 11/76 eottations 11/77 Date 8oard Approval 12/06/77 Date Siging 12/30/77 Datt afteetivenese 03/30/78 11/28/78 Clostg Date 06/30/83 06/30/84 mission DATA

Date mo. of Specialteationm Perforance Types of mission (Go./yr) P ons K ~ epreseanted /b Rating /e Trend/d Probles/e d*ntifieation 01/75 2 7 Preparation - - Apprattal 01/77 4 50 l uS(2),FA total

Supervitson 1 04/78 2 6 L8,6 2 3 K Supervicion 2 09/78 1 2 L8 2 1 N Supervision 3 04/79 2 4 LS(2) 2 1 - Supervision 4 09179 t 1 A 2 3 N,T supervision 5 03/80 2 4 &,A 2 2 1,T Superuision 6 10/80 2 6 A.£ 2 2 N,t,0 Supervision 7 06/81 2 4 &,1 - - Supervialo. 8 08/81 2 6 &,A 3 2 ",T,P Supervision 9 02/82 1 2 4 3 2 N,T,P Supervision 10 10/82 1 2 A 2 1 X,0 Supervigton It 01/83 1 3 A 2 1 0 Supervtison 12 01/84 1 - 2 2 1 0

Total 100 OTEgRPR~C 0ATA Oorrovert The Republic of Sotswanc saccuting Apecy Ministry of Agricultur,

Preceeding Project lama (irst) Livesoek Developmnt Project . Credit Number 323-8T Credt Ao*nt 1.65 Cradit Agreesant Date 06/130/72

Follov on Project wama fattonal Land Hanagenant and ,ivestock Projeet (#LNLp) f Loan Mwaber 2366-T 1.as Asoont O<~ 8 illion) 10.7 Loan Agraesot Date 12/05/85

as of Joly 198 . A • Agricultorist, a - onoUist. L9 - Uvestock Specialist, få Flinancial AMalyst c 1 • Problem-free or oionr problam; 2 - oderate problas; and 3 • Najor probles. 1. Improving; 2 - Stationary; 3 - Datrorattag. F financial; 1 • menagerial; t - technteal; P • political; 0 - othere. PPAR, Botsvan - Firat Livectotk Devalopent Projecc

PROJECT PERFORMANCE AUDIT REPORT

BOTSWANA

SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

EVALUATION SUMMARY

Introduction

The project, which was the second of three through which the Bank is assisting the Government of Botswana to develop its traditional livestock sector, was designed specifically to support a new government land reform initiative, the Tribal Grazing Land Policy (TGLP). It followed the First Livestock Development Project (Credit 325-BT), approved in June 1972 for US$1.65 million which was prinaipally a land reform effort in remote unutilized areas providing funds for development of cattle ranches, sheep farms, management services, borehole maintenance and stock routes. The project was a useful learning process but was concluded to have a negative ERR at completion.

Objectives

The Second Project attempted to avoid the latter's problems (poor management, logistical difficulties, weak accounting, natural hazards) but, through its support to TGLP, had as its maln objectives the continuing development of leasehold ranching, the application of new technologies to help large and small livestock producers, and the improvement of marketing infrastructure and manpower resources. Total project costs were estimated at US$13.4 million of which the Bank was to finance US$6.5 million (53%). The project was cofinanced by the European Development Fund (US$0.3 million) and ILCA (US$1.26 million). In the event, the cost was some US$8.8 million, of which the Bank financed US$6.02 million (68%), US$0.48 million being cancelled after completion.

Implementation

There was some seven months delay in effectiveness from March to November 1978 which foreshadowed subsequent slippage due to legal, adminis- trative and organizational difficulties associated with TGLP to which project progress was directly linked. In fact, the complexity of TGLP which was formalized only in 1975 and itself depended on project support, was under- estimated. The identification of "empty" land, its demarcation, allocation and legal status were all complicated issues requiring coordinated actions from several diverse Government groups. There were further delays over preparation of ranch plan formats and NDB lending procedures, which were complicated in turn by the traditional disposition of the new leasees, provision of collateral, and leasees' attitudes to project objectives. Despite the many difficulties, satisfactory progress was made overall, but it -v - was slow and painstaking. The ranching effort was implemented but on a reduced scale and with less than ideal collaboration from ranchere, the majority of whom continued with traditional ways on their ranches. Attempts to foster communal grasing units (CGUs) and provide small-scale credit to communal management associations were constrained by similar problems and by reported parallel funds from other sources on easier terms. On the other hand, formation of a new credit division in NDB, decentralization of NDB activities and training of its staff proceeded highly satisfactorily and was a project highlight. Overall, implementation of the improved marketing infrastructure component including trek route development, improved animal handling and watering facilities on the routes and at selected railway sites, and smallatock handling facilities, was highly satisfactorily and well in advance of appraisal expectations.

Technical services development covprising ranch management training, M&E of project progress, research and experimental studies, was variable but satisfactory overall. Identification and preparation of a follow-up project was provided for under the project and succesafully accomplished.

Results

Despite a slow startup the project later made up ground and with a one-year extension achieved most of its physical, institutional and technical aesiatance goals. There were shortfalls in the development of CGUs and smallholder credit but these were more than offset in the audit's view oy the fact that progress under the NDB support component and trek route development were some well over appraisal projections. These were significant achieve- ments and, when taken with the successes of the other marketing infrastruc- ture efforts, the positive M&E results and the successful training of some 150 ranch managers, they provide a significant resource in support of ongoing work under the phase III project.

A disappointing project feature at audit was the difficulty of demonstrating quantifiable direct benefits from the ranching component. There was a shortfall in ranches developed (78 actual versus 100 expected), but more important was the apparent failure of the new leasees to develop their ranches and apply new management and technology as was hoped under TGLP concepts. Disappointingly, ranchers are farming their land largely along traditional lines with little inclination to reside on their land, to employ the new class of ranch manager trained under the project (some 150 in all), to apply better management and newer technology, or to market animals seasonally. They switch animals between their ranches and communal areas, are largely uncooperative with professional government workers and NDB staff, and keep few if any records. Where records are given to NDB or DAR they do not reconcile and do not reflect the on-ranch situation in those cases where that might be able to be checked. Under the circumstances it is difficult to monitor production trends. There is an impression among veterinary and extension staff that weaning rates may be up slightly and mortality down marginally on ranches, but there are no data to support it. For these reasons the audit was unable to substantiate the project's direct production impact. - vi -

Sustainability

The project's physical, institutional and technical assistance achievements are considered sustainable by the audit, particularly given government's serious commitment to TGLP and the presence of a third-phase follow-up project. Sustainability of the project's productive elements, how- ever, are conjectural in the audit's view given the observed complex on-ranch situation and ranchers' attitudes* This will only become clearer as TGLP develops with the support of the phase-III project (the National Management and Livestock Project (NLMLP) for which loan 2566-BT in the amount of US$10.7 illion was approved in June 1985) and with further time and inputs. Findings and Lessons

The major lessons and findings arising from this project expetience are:

- the project's physical, institutional and technical assistance objectives were satisfactorily attained overall (PCR, para. 9.04; PPAM, parse. 6-12, 16);

- there were noteworthy project successes in NDB development, and implementation of the training, technical assistance and marketing components, particularly the trek-route orogram (PCR, parase. 3.12-3.49; PPAM, paras. 14, 15);

- the ranching component fell short and direct benefits from it are conjectural and are considered low at best by the audit (PCR, paras. 3.03, 3.04, 5.03-5.07, 9.05, 9.06; PPAM, paras. 14, 16, 17);

- the project's physical, institutional and technical assistance components are considered sustainable by the audit but, given the current complex on-ranch situation and the lack of reliable data, sustainability of the project's productive elements must be considered questionable at this point (PPAM, paras. 18-19); and

- the experience of a number of livestock projects in Western Africa (Senegal and Zaire), which moved away from "ranches" and concen- trated on group development of traditional pastoralists, might be relevant (PPAM, para. 29). - 1 -

PROJECT PERFORMANCE AUDIT MEMORANDUM

BOTSWANA

SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

I. PROJECT SUMMARY

Project Description

1. Loan 1497-BT was the second Bank-assisted project in Botswana's livestock subsector. It was a follow-up to Credit 325-BT (the First Livestock Development Project) which was approved in June 1972 for US$1.65 million, became effective in April 1973, and closed in April 1980 with an undisbuarsed balance of US$50,000, which was cancelled in March 1981. The First Project, which was cofinanced with SIDA, was principally a land development effort in remote unutilized areas and provided funds for develop- ing cattle breeding and fattening ranches, sheep farms, and for management services, borehole maintenance and stock routes. The PCR and auditl/ both reported serious implementation problems arising from poor management, logistical difficulties, non-acceptance by participants, weak accounting, cost overruns, fire hazards and on-ranch damage by migrating animals. The First Project's estimated ERR at appraisal was 21% but in the event was negative at project completion.

2. The Second Project (Loan 1497-BT) was identified, prepared and appraised in the latter half of the First Project's lifetime (full details in PCR, paras. 2.01-2.12). It was designed specifically to avoid the latter project's problems and to support a recently developed government initiative -the Tribal Grazing Land Policy (TGLP)-aimed at developing the country's livestock potential by reducing overstocking and range deterioration through reform of land tenure. As a central part of TGLP the project aimed to support:2/

(a) development by private farmers of about 100 ranches on land designated "commercial" under TGLP:

(b) development by the Animal Production Research Unit (APRU) of some 14 Communal Grazing Units (CGUs)-12 CGUs and 2 control units-

I/ Project Performance Audit Report, Botswana - First Livestock Development Project (Credit 325-BT), OED Report No. 3989, dated June 22, 1982.

2/ Project description as per SAR (para. 4.01); for more details see PCR, para. 2.13 which provides details on project components along the lines of SAR chapter IV. SAR: Botswana - Second Livestock Developnent Project Appraisal Report, Report No. 1630-BT dated October 31, 1977. - 2 -

aimed to test different grazing systems on land designated 'communal" under TGLP;

(c) small-scale loans for livestock developments by communal management associations on communal land;

(d) support to a new agricultural credit division of the Natioital Development Bank (NDB);

(e) development of livestock marketing infrastructure, including trek routes, railway handling facilities and small stock holding grounds; and

(f) technical services, including studies and experimental work, establishment of a ranch training center, monitoriAg and evaluation of project progress, and preparation of future projects.

3. A Project Steering Committee (PSC) chaired by the Deputy Permanent Secretary, Ministry of Agriculture (MOA), and comprising .e Director of Economic Affairs, Ministry of Finance and Development Planning (MFDP), and the General Manger, NDB, was assigned overall responsibility for policy direction and project coordination. For day-to-day coordination a Livestock Project Coordinating Committee was created, comprising all officials with rk;sponsibilities for project implementation, chaired by a MOA Livestock Project Development Officer who was also secretary of the PSC. The latter officer would: "... organize the work of the Committees, ensure that decisions are implemented, monitor the provision of projects funds, prepare periodic reports on project Progress, and oversee preparation and audit of accounts" (SAR, para. 5.01). He was a de facto project manager, but was not so designated, the title of Project Coordinator (PC) being preferred locally.

4. Total project costs were estimated at US$ 12.18 million equivalent, net of taxes, of which the Bank agreed to finance US$ 6.5 million (53%), including 82% of foreign costs and 11% of local costs. The remaining project costs were to be shared as follows:

US$ Million Z

European Development Fund (EDP) 0.30 2 International Livestock Center for Africa (ILCA) 1.26 10 Ranchers 0.81 7 Government of Botswana 3.31 28

The actual final cost of the project was US$ 8.77 million equivalent. Bank financing totaled US$6.02 million (68%), US$ 0.48 million being cancelled in March, 1985 after the project's close. -3-

Project Implementation3;

5. The project experienced seven months initial slippage from March to November 1978 due to Government delays in satisfying effectiveness conditions. Even after the project was effective, implementation continued to be protracted tdcause of numerous legal, administrative and organizational difficulties surroundiag the effective functioning of TGLP to which project progress was directly tied. The difficulties included slow land use planning decisions, delayed identification, demarcation and allocation of commercial ranching areas, the legal status of demarcated land, ranch plan formats, and NDB lending procedures for ranch development. A shift in TGLP policy also occurred as it became increasingly clear to government staff that many areas of the country considered to be unoccupied and thus thought suitable for demarcation for commercial use, were in fact occupied by substantial human and livestock populations. The task of identifying "empty" areas that could be zoned for commercial ranchers to move their cattle to and subsequently develop while at the same time reduce stocking pressures in communal areas proved much more difficult than was foreseen. For these reasons implementa- tion of both commercial ranching and the CGUs was seriously constrained and, in the event, only 78 ranches and one active CGU and a control CGU were established in the course of the project. Small-scale livestock loons to communal management associations were also seriously constrained by general project problems and, reportedly, by the availability of funds from alternative sources on more favorable terms, which precluded development of the component in its entirety.

6. On a more positive note, progress with the establishment of an agricultural credit division within NDB proceeded largely as planned and was a highly successful project outcome. FAO provided a manager for the division who made a key contribution to project progress. During 1978/79, six new NDB regional branches were established which by 1981 were mostly housed in new offices provided under the project. Because of increasing loan activity over a wider geographical area, NDB further decentralized its operations in the early 1980s through the establishment of 12 district credit offices and the introduction of a decentralized loan approval scheme providing greater discretion for regional managers. All of these moves were designed to reduce delays between subloan application and approval and generally to improve services to clients, which they did very successfully. At the same time, NDB implemented an extensive staff training program to provide the manpower for its new branch network. By late 1983, some 62 professional staff had received intensive in-service coursework and 14 more had completed courses abroad. An additional 28 clerical staff participated in training courses appropriate to their responsibilities (PCR, paras. 3.12-3.13). NDB's training exercise was well planned and executed with the result that by project close the new division was beaded by a credit anager with degree training and the six regional and 12 district offices were all operating with appropriate managerial and support staff. The generally high standard of NDB operations and commitment and competence of its staff was clearly evident at audit.

3/ For more extensive details see PCR Chapter III. 7. Implementation of the improved marketing infrastructure component was also highly satisfactory overall. It was a large and complex undertaking involving:

(a) development of 28 stock holding and watering facilities on three existing and 12 new trek routes;

(b) development of holding grounds, yards, loading ramps and water facilities at 15 selected railway sites;

(c) provision of 16 handling and loading facilities on trek routes to provide for lorry pickup and trucking of cattle; and

(d) provision of five smallstock holding grounds with dipping and handling facilities to assist cooperatives in smallstock marketing.

Complications including shortage of land (particularly in the east), difficulty in deciding relative priorities of different activities, foot and mouth disease outbreaks (impeding livestock movement in affected areas) and problems in locating suitable water supplies (many boreholes were dry or salty) slowed initial implementation. The magnitude, logistics and complexity of developing trek routes in isolated sites was initially under- estimated and the lack of a comprehensive plan with priorities and an imple- mentation schedule surfaced early on as a component constraint. MOA was thus charged through its Department of Animal Health (DAR) and Department of Water Affairs (DWA) with developing a National Trek Route Policy (NTRP) which it successfully did in 1979/80, after which implementation progressed satisfac- torily in both trek route development and all associated aspects of the improved marketing infrastructure component (further detailed in PPAM, paras. 14-16).

8. Implementation of the technical services component comprising establishment of a Ranch Management Training Center (RMTC), M&E of project progress and research and experimental studies was variable but satisfactory overall. The project also provided considerable manpower input into the identification and preparation of a follow-up phase, which was successfully accomplished.4/

9. The Ranch Management Training Center (RMTC), which was funded by EDF, was successfully established in 1978/79. It commenced its first training course in 1979 and continued thereafter with annual intakes of selected students who followed the prescribed two-year practically-oriented course as planned. Some 150 ranch maiiagers were trained by RMTC in the project period but the effectiveness of the program tended to be reduced by the fact that, after completing their studies, a majority of graduates had to

4/ The third-phase project entitled the National Land Management and Livestock Project (NLMLP) was approved by the Board in June 1985 (Loan 2566-BT) and became effective April 10, 1986. SAR, Botswana - National Land Management and Livestock Project, Report No. 5471-BT dated April 29, 1985 (see also Annex 3). - 5 - seek employment elsewhere as the terms and conditions on the newly-formed ranches were by and large unattractive, ranchers preferring to continue their traditional ways using lowly-paid cattle hands. This was considered partly due to the concept of professional ranch management being novel and untried and to the fact that the financial problems faced by many of the new ranches (drought, disease, water problems, wildlife intrusions) resulted in weak cash flows which made financing of the new class of manager difficult. This is a problem which under the circumstances will take time to correct (further discussed in PPAM, para. 24).

10. Monitoring and evaluation under the project took essentially three forms:

(a) participation by the International Livestock Center for Africa (ILCA) in reviewing project implementation;

(b) monitoring by the Rural Sociology Unit (RSU) of the project's impact on TGLP implementation; and

(c) more recent monitoring and assessment by the Ranch Extension Unit (REU) of changes occurring on the newly-formed ranches.

The most comprehensive of these was the ILCA effort as it ran for most of the project period (through December, 1983) and was carried forward by two highly-trained consultants and other miscellaneous short-term ILCA staff who, along with local counterparts, made valuable inputs in the areas of seasonal livestock and wildlife movements, and geological and vegetal interrelations and their effects on livestock nutrition and productive performance. ILCA findings also resulted in valuable practical recommendations on fencing layout (particularly in relation to variations in vegetation types), in controlled use of water, and in selecting and studying human communities for extension work (PCR, paras. 3.45-3.40).

11. The ME efforts of the RSU started effectively in mid-1980 and that of REU in 1983/84 (i.e. later in the project period). They produced valuable insights into various aspects of TGLP and the progress of the newly-formed ranches. Their work will continue under the ongoing third-phase project. The most striking observations on the new ranches included the almost universal absenteeism of owners and their choice to continue for the most part along traditional lines, frustrating efforts by REU to introduce new, improved methods and thereby improve overall ranch productivity. There were tentative indications that weaning rates may be increasing and mortality decreasing on some ranches, but there were no data to substantiate them. Such gains, even if they existed, were offset by observations that ranches generally were seriously overstocked and that the new lessees were switching animals between their ranches and communal grazing areas depending on grazing conditions, which tended to lower the reliability of any on-ranch production statistics and to frustrate the most basic tenets of TGLP. Furthermore, leasees were reportedly not sympathetic to the concepts of reducing stock numbers nor controlling the intensity and frequency of grazing which were essential measures to halt degradation/desertification and encourage rehabilitation. - 6 -

12. Implementation of the above M&E efforts were supported by five successful stocking rate and grazing system trials undertaken by APRU aimed at evolving management strategies to minimize overgrazing, range degredation and desertification. In the relatively short timeframe of the project a valuable base was laid from which to launch future, more-concerted efforts but, unfortunately, little by way of practical significance for immediate use by producers was achieved. These longer-term studies will be continued under NLMLP (Loan 2566-BT). APRU also undertook responsibility for implementation of the CGU component which, with the exception of one control unit and one ongoing CGU (which continues somewhat precariously), was largely unsuccessful for both technical and socio-economic reasons.

Project Results

13. The project was designed to support a new government initiative, TGLP, seeking to rationalize the utilization of Botswana's extensive range- land resources which were being exploited communally and haphazardly to the detriment of both the range and the livestock subsisting on it. TGLP was instituted by Government in 1975 as an innovative development approach essentially to improve the management and productivity of the rangeland (much of which was becoming degraded and desertified through overstocking) through the introduction of leasehold ranching on what were previously tribal (communal) lands. It was a novel, bold and long-term undertaking and the achievements of the project, which was designed principally to implement the new policy, might usefully be viewed against this background. Project achievements might also be seen in the context that Livestock II was a follow-up to an earlier Bank-assisted effort to support the livestock subsector in Botswana which at audit in March 1982 was concluded to be an "unhappy experience" and to have a negative ERR.

14. Given these contexts, and taking account of the exceptionally difficult agricultural environment in which the project was being imple- mented, and the fact that local staff were relatively scarce and largely inexperienced, the physical achievements under the project were surprisingly good. Actual achievements against appraisal expectations are summarized in Annexes 1 and 2 which are reproduced with minor editing from the 1983/84 (final) project annual progress report prepared by the PC. There was a shortfall in the expected number of ranches developed (78 versus 100) and in the formation of CGUs (2 versus 14). In addition, small-scale loans to communal management associations on communal lands were not made because of organizational difficulties within the associations which precluded demand for such loans. On the other hand, trek route development exceeded appraisal projections by better than 100%, notwithstanding that the choice and siting of routes had to be changed in line with implementation experience to make them fully effective.5 / Again (PPAM, para, 6), the establishment and development of six NDB regional and 12 district credit offices together with the successful training of their full complement of staff were each 100% above appraisal expectations. Technical services were implemented

5/ There were 57 stock holding and watering facilities developed in all in the project period compared with 28 anticipated at appraisal. - 7 - successfully largely in line with appraisal planning, including the establishment of the RMTC (and the implementation of accompanying ranch personnel training), monitoring and evaluation of project and TGLP progress, and stocking rate/grazing system and related trials and studies. The technical services component overall produced a relatively large complement of well trained ranch managers (some 150 in all) together with a volume of qualitative and quantitative information providing an invaluable base which not only assisted the later stages of the Second Livestock Project itself but formed the foundation for the identification and preparation of the third phase project -- NLMLP (which was a further objective of Livestock II which was successfully achieved).

15. Along with the highly successful trek route program, the results of other aspects of the improved livestock marketing component were fully satisfactory and up to appraisal targets. These included development of the targeted number of 15 railway handling facilities (holding pens, watering potats and loading ramps) and the smallstock holding ground facilities. It was clear at audit that the railway handling facilities were being extensively used and thus serving the purpose for which they were provided. Unfortunately, the smallstock holding ground facilities were apparently not regarded by intended beneficiaries as being particularly useful and showed little evidence of use. In two of the five smallstock facilities reviewed by the audit mission there were obvious design and construction flaws and both sets of dipping facilities lacked a permanent water supply, thus greatly detracting from their usefulness and impact.

16. In summary, the physical, institutional and technical assistance achievements of Livestock II were on balance satisfactory, such shortfalls as there were being counter-balanced by useful gains in other areas. These were, however, all indirect project benefits. Insofar as direct benefits were concerned--and these were anticipated at appraisal to include increased production and improved quality of beef, increased offtake of cattle and smallstock, and reduced transport losses--project results were much less clear to the audit. At appraisal the ERR of the project was estimated at 15%, based on quantifiable benefits derived from:

(a) incremental sales and income from ranches due to their superior management over traditional cattle camps;6 / and

(b) estimated lower weight losses in animals in transit occasioned by the project's improved marketing infrastructure.

In its estimation of benefits the appraisal mission emphasized that "...the project is designed as a pilot effort of a broad and long-term development program. Benefits that would result from investment in this initial program are subject to considerable uncertainty. Development should be viewed in part as experimental in nature, designed to contribute experience that would assist in design of future, almost certainly larger, projects" (SAR, Annex 16, para. 2).

6/ Known locally as "cattle posts". - 8 -

17. The appraisal mission's caution in mid 1977 was, in the audit mission's view, commendable. It was also prophetic in that in November 1985, with over eight years of project experience and hindsight, it is difficult for the audit to be certain that there are significant direct benefits resulting from the project. This arises because of the extraordinarily complex circumstances within TGLP and the difficulty of identifying and quantifying differences in production, if in fact they exist, between animals farmed on the new ranches and those grazing under communal management on common areas. Government recognizes these problems all too well (PCR, paras. 9.05-9.06) and in its PCR did not recalculate a rate of return or even conjecture on productive benefits accruing from the project's productive investments. The audit identifies with this approach for the several reasons expanded in Section II (PPAM, paras. 20-28).

Sustainabil ity

18. The audit considers the physical, institutional and technical assistance achievements of the project highly sustainable. This is particu- larly so given that the project was a principal vehicle for the implementa- tion of Government's innovative, long-term TGLP land reform efforts aimed at more rational utilization of the country's extensive rangeland resources and of the livestock which graze on them. TGLP is an evolving system aimed at basic restructuring of traditional land tenure and animal grazing patterns through the development of ranching, the application of new technologies and the promotion of more efficient livestock marketing. Government is deeply committed to implementation of the program and regards TGLP as a central plank of its long-term development plans for agricultural sector develop- ment. The further development of TGLP vill be carried forward in the immediate future by the third phase Bank-assisted NLMLP, which became effective April 10, 1986.

19. As noted, the audit views the status of the project's productive elements as highly conjectural which cannot be demonstrated or quantified at this juncture. This situation is expected to clarify in the course of the implementation of NLMLP, which has several elements specifically designed for this purpose. Until this clarification is made, the direct productive benefits of Livestock II should, on the basis of the audit's field observa- tions, be considered marginal at best. Under the circumstances, the question of the sustainability of such benefits cannot be decided at this juncture.

II. ASSESSMENT OF PROJECT IMPACT AND RATIONALE

20. TGLP represents a serious long term effort introduced in 1975 by the Government of Botswana to reduce overstocking, to improve range produc- tivity by the introduction of new -:echnology and to increase livestock off- take in the country's extensive rangelands which comprise some 90% of its land resources. The cornerstone of TGLP is the introduction of commercial ranching by offering exclusive use of designated leasehold areas to the country's larger herd owners to enable them to develop the areas on an individual or group basis and thereby reduce the stocking pressure on - 9 - communal areas which would gradually be improved through group action for use mainly by smaller, subsistence-type livestock owners. The exercise involves the designation of certain low population density areas of the country for commercial use and subsequent demarcation and allotment of the individual ranches to leasees for development through the application of improved management and newer technology. The approach was predicated on the assumption that there was no feasible method to rehabilitate and develop the country's extensive rangelands so long as they remained overstocked and subject to opportunistic grazing under traditional communal management. In the 1960s and 70s the need to confront the problem became more and more urgent as stocking and grazing pressures intensified and as sizeable areas of the country's vital rangeland resource became overstocked and degraded through uncontrolled borehole development and unsustainable increases in livestock and wildlife. The Second Livestock Project was designed specifically to assist with the launching of this important effort.

21. As noted, the project's physical, institutional, training and technical assistance components developed satisfactorily, notwithstanding some early slippage. The progress of certain activities far exceeded appraisal expectations, of which trek route and marketing infrastructure development and support to NDB were examples. Other components such as the RMTC training activities, studies and research and M&E efforts were fully up to appraisal expectations. In the audit's view these highly positive developments more than offset the disappointing CGU and smallholder livestock loan components, the impact of which was minimal, an outcome perhaps not wholly unexpected in view of their pilot nature.

22. The project's most significant shortfall, particularly in the context of TGLP development, was that of its ranching component. Development of some 100 ranches was projected at appraisal but in the event only 78 were actually developed in the project period. The shortfall of 22 is one aspect but, much more important, is the disappointing on-farm development and the lack at audit of improved production parameters on the 78 ranches developed under the project. This is both a major project and a TGLP issue since it raises several key questions concerning the basic rationale of TGLP and its effectiveness as an instrument in rangeland and livestock development.

23. Government and Bank hopes were that assignment of exclusive grazing rights to individuals or groups in the designated commercial ranching areas would be a catalyst for owners to improve their management, apply new technology and thereby increase both range and livestock productivity. Improvements were not expected overnight and the appraisal report did not minimize the risk element. However, the project's estimated ERR of 15%, the benefits for which were to come largely from improved offtake from the ranches, provides more than a hint that something fairly tangible was expected.

24. The audit was unable to determine with any degree of certainty that any direct productive benefits were accruing from the ranch component. There were suggestions that weaning rates may be increasing slightly and mortality may be decreasing marginally. These, however, were merely impressions of extension and veterinary officers, without any reliable data to support - 10 -

them. In fact, the lack of reliable data concerning the current on-ranch situation is a major project grey area. In theory, each 6,400 ha ranch has the capability of carrying some 400-450 livestock units, but monitoring this is impossible given the local situation. Individual ranchers are often uncooperative and misleading. Contrary to expectations, the ranches are run on largely traditional lines (see also PCR, para. 5.06) and staffed almost without exception by illiterate herd boys 7 / who have no information to divulge to professional government staff (or, if they have, refuse to do so). So far as the audit could ascertain, there are no written livestock records available on any of the ranches. Government staff report that figures given by owners on NDB credit applications do not reconcile with veterinary and vaccination records, and these data in turn are found to be different from the on-ranch inventory, if and when it is possible to obtain. Further complicating the situation, ranch leasees switch cattle back and forth between their ranches and communal areas, refusing to relinquish their access to the communal areas, thus enjoying dual rights, which largely defeats basic TGLP objectives. At audit, the several ranches visited by the mission were as seriously overgrazed as adjacent communal areas. It was in November at the end of the dry season and it was exceptionally dry and available grazing was at its lowest. Had the ranches been better managed it would have been an appropriate time to observe differences. To the audit there were no discernible differences between the ranchts and the traditional range areas in either feed availability or livestock condition. Again, with no reliable data to substantiate the suggested higher weaning percentages and lower mortality, the audit feels, based on what it saw and on local professional staff opinions, that any productivity differences between the ranches and the traditional areas are at this stage extremely questionable. In the same context, while trek route and livestock handling facilities were successfully developed under the project, there were no data to substitute reduction in weight losses as a consequence of the developments. These may come in time but, so far, they appear not to exist. It is for these reasons that the audit concluded that the project's direct benefits must be considered conjectural and their sustainability (assuming they exist) is an open question which cannot be decided without reliable data (PPAM, para. 19). This is disappointing but perhaps not wholly unexpected given the innovative, uncertain and long-term nature of TGLP, which the project was designed to benefit, and the risk emphasized at appraisal (PPAM, para. 16). In terms of its physical, institutional, training and technical assistance

7/ Notwithstanding the success of the RMTC component in graduating some 150 trained ranch managers, few of these have been employed for the purposes for which they were trained. Explanations vary for this situation. Some government staff feel that the ranchers are too conservative to employ professionally-trained managers, and will remain so far some time to come. Others maintain that the ranchers simply do not want educated staff running their ranches and knowing too much of ranch affairs. Still others are of the opinion that ranch cash flows cannot sustain the wages asked by the new graduates (herd boys receive only a reported Pula 5-10 (US$2.50-5) per month, plus some milk and other perquisites). The true explanation was not clear to the audit but probably lies somewhere between these various views. - 11 - achievements, the project has on balance been a highly successful learning experience and a valuable stepping stone in TGLP development. Such shortfalls as there were have been faced frankly by both the government and the Bank and the lessons learned appear to have been fully taken into account In the phase-three project which Is expected shortly to be effective and which in fact contains specific components to address problems contributing to phase-two project shortfalls.

25. A review of the phase-three documents by the audit and discussion of that project's design and objectives with senior M0A staff at the mission's wrap-up meeting in Gaborone, provided strong reassurance that Government was keenly aware that TGLP was a complicated, long-term exercise in a vital land reform area bearing directly on the country's economic future. There may have been some in Government who earlier considered that the exercise could be accomplished easily, quickly and without complication. Even some Bank staff may have felt similarly. The experience of the Second Project has, however, left no doubt in anyone's mind of the difficulties and of the fact that it may require 20-30 years (i.e. a full generation) to change the current situation significantly. This is already clear in government's mind and hopefully also no doubts remain in Bank thinking on the matter.

26. The two completed Botswana livestock projects contain a number of elements common to many projects in the East and Southern Africa region, namely overly ambitious objectives which were expected to be realized in an unrealistically short timeframe and an ERR appraisal estimate which, with hindsight, was not really attainable. 8 / Both projects would undoubtedly have appeared much more successful had their objectives been less ambitious, their timeframe more realistic and their ERR more modest. Both were novel, groundbreaking projects with little real possibility of achieving significant productive benefits in a five-year timeframe. The audit considers that it would have been better to recognize this from the outset and to plan a 20-30 year development program appropriately timesliced to reflect the need initially to refine policies, develop institutions, train staff and to under- take research and develop packages and systems. It would certainly have been better to set lower production targets and to delay all but essential on-ranch investments. Given that leasees decided to continue their tradi- tional ways, much of the fencing program, for example, might have been post- poned until a firmer production base and stronger cash flow were apparent on individual ranches. Again, the whole rationale for high-cost traditional fencing in the local situation might appropriately have been more carefully researched from the outset, particularly given the uneven results from fencing of ranches in the First Project. It would also have perhaps been more appropriate in the local context had the project been justified more on arresting a clearly declining range situation in the first instance rather than on arresting and reversing it, as was implied from the project's relatively high expected direct benefit stream and its projected 15% ERR.

8/ The Smallholder Dimension of Livestock Development: A Review of Bank Experience, OED Report No. 5979, dated December 18, 1985 (paras. 10, 11, 14, 34 and 35). - 12 -

27. Bank staff argue that such an approach would probably not have been agreed by the Board--also that an ERR above 10% Is part of an agricultural project package, without which the project would neither have been approved nor launched.v/ If this is in fact so, then possibly the Bank's approach to long-term sector development requires further thought, particularly where difficult socio-economic, political and structural changes in a highly traditional and relatively undeveloped sector underly project objectives. If not, then perhaps the Bank's and its borrowers' interests would be better served if staff were to calculate ERRs more conservatively in cases where direct benefits are dependent on complicated and delicate sectoral structure changes and the likelihood is low of those benefits accruing in the project's development timeframe. A statement concerning the project's high risk is useful but is not necessarily an adequate substitute for a low ERR if that outcome is a high probability from th% start, as appeared to be the case in in both the phase-I and phase-II livestock projects in Botswana.

28. In the rational development of Botswana's large and important arid/semi-arid land resource, a start had to be made from almost ground-zero in planning strategy, formulating policies, developing projects, and in building institutions, technology, skills, commitment and confidence. To achieve in all of these areas as the project did, and to expect a significant parallel production impact was expecting too much in the audit's view, particularly in the local harsh and unpredictable climatic context, and in such a short timeframe. That the project achieved what it did was commend- able and sufficient justification for Bank and Government support of it. It formed a valuable stepping stone of several required to develop the country's range potential fully. Direct production benefits will not accrue any more easily nor quickly in the immediate future than in the past, but then they will never accrue at all unless the necessary developmental infrastructure is put in place as a prerequisite. It is an essential overhead which must be met if long-term development is to be successfully achieved. In developing such infrastructure the project provided an important impetus to livestock subsectoral development and an essential base from which future more substan- tive production-oriented efforts can be launched.

2.9. However, experience with other livestock projects might be relevant for the Botswana livestock development strategy. A number of livestock projects in Western Africa have moved away from "ranches" and targeted their efforts to traditional pastoralists by organizing them in potentially self sustaining groups. So far two such projects in Senegal and Zaire have been rather successful.10/

9/ This perception is not completely born out by facts. The early rural development projects in Malawi were justified as part of longer term programs and some had rates of return lower than 10%. See for example PPAR, Malawi-Lilongwe Land Development Program Phase I (Credit 113-MAI), OED Report 751-MAI dated May 23, 1975.

10/ PPAR, Senegal - Eastern Senegal Livestock Development Project (Credit 633-SE), OED Report under preparation and PPAR, Zaire - Ituri Livestock Development Project (Credit 697-ZR), OED Report under preparation. - 13 - Annex 1 PROJECT PERFORMANCE AUDIT MEMORANDUM Page 1 BOTSNANA - SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

GEOGRAPHICAL DISTRIBUTION OF PROJECT SITES

•••••o or

hux.

LEGEND .....Ronch Dovelopment (78) .*.- ..... Trok Route Developmen (57) s ...... Communal Grazing Units (2) . T ...... ND111 Regional Finonce Office$ ( 6) ...... Stocking Rote Trials(5) .....- . Roilwoy handling Pacilities(16) ...... N DS pistrict Credit offices(12) ...... $m~Ii Stock Holding Gröunds (5) ...... Ronc- Management Tröining Centre - 14 - BOTSWANA - SECOND LIVESTOCK DEVELOPMENT PROJECT Annex 2 (LOAN 1479-BT) Page 1 PROJECT ACHIEVMENTS'

Activity Proposed Completed

1. Ranch Development 100 (78) Ngwaketse: Semane 7 Keng 5

Central: Lepasha * 5 Nata 4

Kweneng: 9

Ngamiland: Haina- veld - 48

2. Trek Route Develop- 28 lo** (57) ment Francistown/Maun * 12 Maun/Shakawe * 10 Rakops/Makoba . 4 Gantsi/Lobatse 31

3. Communal Grazing 14 (2)

Makhi II I Ntimbale 1

4. NDB - Agricultural Credit Division (i) Regional Finance Offices 3 (6)

Maun/Francistown/Selebi- Pikwe/Serowe/Gaborone/ Kanye (ii)District f Credit Offices 6 (12)

Tutume/Mmadinare/Bobonong/ Mahalapye/Mochudi/Moshupa/ /Ramotswa/ Good-Hope/Mmathethe/ Tsabong/Ghanzi

If From 1983/84 (final) project annual progress report prepared by the Project Coordinator. - 15 'Annex 2 Page 2

Activity Proposed Completed

5. Stocking Rate 5 (5) Trials Tsetseku/Lesego/ Seleka/ Masiat ilodi/Xanagas

6. Railway Handling 15 (16) Facilities I Notwane/Sebele/Pilane/ Artesia/Dibete/Phala Road/Mahalapye/Radisele/ Palapye/Serule/Selebi- Pikwe/Shashe/Francis- town/Bosoli/Tehesebe/ Ramokgwebana.

7. Smallstock Holding Grounds 5 (5) Semolale-Kobojango/ Molalatau/Bobonong/ Tsetsebjwe/Tonota.

8. Ranch Management Trai1ing Centre 1 (1)

Ramatlabama

9. Small Scale Communal Grazing Un-speci- Schemes fied None - 16- Annex 3 Page 1 PROJECT PERFORMANCE AUDIT MEMORANDUM BOTSWANA - SECOND LIVESTOCK DEVELOPMENT PROJECT (LOAN 1497-BT)

FOLLOW-UP PROJECT-

Project Rationale and Objectives

2.01 As a major source of foreign exchange, the growth of the cattle industry is crucial to Botswana's overall development. It also directly effects the welfare of a large proportion of the population. Although over the last '.0 years, commercial sales from the cattle industry have expanded to fully utilize the country's abattoir capacity, there is growing evidence of serious overgrazing and range degradation, which will threaten the long term viability of the industry. It is apparent that conservation of the grazing resource can only be achievid if overgrazing on the tribal lands, which presently account for 65% of total land area, can be reversed. To achieve this, market-oriented producers (whether individuals or groups) will have to be linked with demarcated grazing areas, and "free" access to remaining, truly communal areas will have to be limited to and controlled by the remaining subsistence or very small scale producers.

2.02 The Government's Tribal Grazing Lands Program (paras 1.30-1.36), a wide ranging land tenure reform program unique in Africa, was designed for this primary purpose. If implemented along the guidelines of the original policy document (para 1.31), TGLP could achieve sustainable output of fodder from the grazing resource, which in turn would lead to an expanded level of livestock production. However, it is evident that TGLP needs to establish a sound planning base before the majority of the country's farmers can be brought into the program. It will also require the strong capability to sort through and settle, in a timely and equitable manner, what in some cases will be difficult adjudication issues.

2.03 Simultaneously, while land tenure changes are taking place, the Government recognises that it will also have to tackle two other important facets of livestock development: the need to assure adequate processing capacity to handle the increased offtake which would come from better range management, and the need to refine the package of fiscal and pricing incentives and disincentives not only to discourage the overgrazing of the range but also stimulate production of the quality of meat its export markets demand. It will be necessary to increase abattoir capacity. The Government is proposing to build a new abattoir (para 1.26) and has submitted to the Bank a timetable for the design, construction and commissioning of the plant. Trek route development will also need to be furthered. With increased marketing capacity overall, there should be less holding of cattle beyond their optimal selling rate, less reduction of weight and quality enroute to the abattoirs and less pressure on grazing lands. However, in the long term, control of the national herd size will require careful balancing of marketing capacity with appropriate fiscal, pricing and land use policies. As to fiscal and pricing incentives, the Government will be embarking on a review of a variety of means (the pricing

1/ Taken verbatim from SAR, Botswana - National Land Management and Livestock Project, Report No. 5471-BT Dated April 29, 1985 (paras. 2.01-2.06). - 17 - Annex 3 Page 2

structure for different grades of meat, tax incentives related to improved range management, cost recovery for services, and so forth) which, at the same time, reinforce producer incentives to adopt improved production and conservation methods; assure reasonable recovery by Government of resources put into the sub-sector, particularly when commercial benefits to producers are clear; avoid benefitting continuation of unimproved livestock production, or allowing this production to enjoy any artificial advantage over the other land uses (cropping, wildlife, etc.).

2.04 To help address the various above needs (except for the abattoir), the Project would strengthen the Government's institutional capability to promote, through both program and policy interventions, more effective utilization of the rangeland resource and further development of improved livestock production system. Direct output from the Project would be: the production of four land-use plans for the Central, Ngamiland, and Southern Districts which have 70% of the total cattle population and 86% of Botswana's livestock holders; assistance and training in the specification and application of improved methods and procedures for implementing the land use plans; refinement of the pricing and taxation policy package, an implementation plan for the package and training of Botswana personnel who would initially be part of the working parties studying these issues; development of guidelines and plans for management of communal grazing areas and funding to assist such initiatives; provision of credit and managerial assistance for developing both individual and group commercial ranches; and construction of trek routes.

2.05 About 1,000 farmers will be directly involved in the proposed production components but the Project's planning activities have been designed to create the basis for the eventual involvement of all 58,000 livestock farmers in the Tribal Grazing Lands Program. The project's concentration on securing an improved national capacity to manage and exploit rationally and equitably its centrally important rangeland resource is the basis for the broad indirect benefits. While technical assistance, training-and administrative development represent a large share of project activities and costs, these are in aggregate both appropriate to the nature of the needs and relatively modest in comparison to the economic importance of the sub-sector and its future sustainability and growth potential. They also reflect the considerable difficulties which can be expected, and the patience which will be needed in securing for the subsistence producer better control over, and development of, the communal resources they depend upon.

2.06 Specifically, the Project would include the following components:

1. Land Use Planning and Management

(a) Coordination of Land-Use Planning and Training The Project would provide a technical specialist to help the Ministry of Local Government and Lands centrally coordinate all land-use planning activities and ensure that the detailed land use plans are consistent with national resource management policies. In addition, it would fund the operation of related training programs and a land use planning training specialist to help develop the required training courses for all land use planning cadres. Annex 3 18- Page 3

(b) District Planning Unite

The Project would support the establishment of four district planning units, which would be responsible for producing detailed land use plans. Funds would be provided for incremental staff salaries, housing, equipment and vehicles.

(c) Land Boards

To assist the physical implementation of land use plans, the Project would strengthen four land boards through provision of support staff, housing and vehicles.

(d) Land Development Division

To increase the professional capability and facilitate replacement of existing expatriate staff in thir technical support unit in the Ministry of Agriculture, both under- graduate and postgraduate training would be provided under the Project. In addition, a consultancy would be funded to establish the relevant methodology for rangeland analysis for the emerging Range Ecology Unit.

(e) Wildlife Department

The Project would assist Government assess the future structural, manpower and budget requirements of the Department of Wildlife and National Parks.

2. Trek Routes

To increase the effectiveness of the main northern and southern trek routes, the Project would upgrade existing infrastructure and provide .a holding ground and new fenced boreholes. In addition, improved supervision and maintainance of all trek routes would be supported through provision of incremental staff and vehicles.

3. Extension

The Project would provide livestock extension specialists to specifically assist groups of small farmers. Salaries would be provided for two livestock specialists, a rural sociologist together with equipment, vehicles and a fund to develop range improvement infrastructure for groups of small farmers.

4. Credit

NDB would receive Project funds on-lent to them by the Government to provide credit to medium and larger farmers for ranch development. The Project would also fund salaries of a modest number of incremental specialized staff and transport required to supervise these new credit operations. Annex 3 - 19 - Page 4

5. ProJct Coordination

A Project Coordination Unit (PCU), responeible for overall Iaplementation and execution of the Project would be established as part of the MOA. The Project would provide for a Project Coordinator, local professional, support staff and transport.

6. Livestock Pricing and Financial Incentive/Disincentive Analysis

Two internationally recruited specialists would be provided to help the Ministries of Agriculture and Finance develop recommendations on a livestock pricing and financial incentive/disincentive strategy aimed to help make that strategy as consistent as possible with the desired type of long-term development of the livestock sub-sector; the specialists would also help the related Ministries establiih an institutional capacity for the on-going review of prices and taxes. ‘ー タ ジ ー

メりコ J-糾 ごご 多ル<k THE WORLr) BANK INTURNATIONAL FINANGE CORPORATION - 21 - OFFICE MEMORANDUM D= iFebruary 25, 1986 TD: Director, Operations Evaluation Department

FM : Eastern and Southern Africa Regional Office

SU&TECT: BOTSWANA - Second Livestock Development Project (Loan 1497-BT) Project Co!gletio;- Report _

1. 1 attach two copies of the Project Completion Report prepared by the Ministry of Agriculture on behalf of the Government of Botswana In terms of Section 4#07 of the Loan Agreement. We have reviewed the PCR and are in broad agreement with its conclusions* However we would like to make two comments, one on the lessons learned and the other on benefits.

2. In pars. 9.07 of the PCR the Government has laid out four conditions that it considers necessary for future successful ranch develop- ment* While we agree with this analysis in respect of livestock extension and research (issues now being addressed by the Bank and FAO) we have reservations with the PCR's prescription for establishing grazing rights and stocking controls# The Presidential Commission on Land Tenure (referred to in para 9,07(a)), was almost entirely concerned with urban land problems. Its only comment concerning ranch development in the tribal lands was that the one man, one ranch (6,000 ha) approach of TGLP could be modified to allow one man two ranches. However, even this recommendation apparently found little support in the Parliament.

3. The PCR looks at grazing rights and stocking control issues in the narrow context of ranch development while we believe they should be addressed as the cornerstone of any strategy for sustained productive use of the country's natural resources. Limiting ranchers access to communal lands is : simplistic solution to the overgrazing problem and ignores the complex livestock ownership and holdings within the national herd. The tribal grazing lands cover a total of about 37 million ha within which 108 million ha (5%) has been subdivided into 300 leasehold ranches. The remaining 35 million ha holds approx 2.4 million cattle in individual herds which range up to holdings of 5000 head. No official grazing controls exist in this area but many farmers own boreholes on cattle posts that gives them a de facto exclusive use of the grazing. A considerable number of larger herd owners have opted out of taking ranches for fear of government controls and possible taxes. EAPSA's focus in the new Bank Land Management and Livestock Project has been to address the central issue of sustained resource use for the industry through a concerted thrust of land use planning and allocation within the original TGLP guidelines. Such a thrust it is hoped will link specific land areas to individuals and groups of farmers and, over time, eradicate the present "tragedy of the 2ommons". In addition, to increase the offtake for stock control the Government has agreed to build additional slaughter capacity to allow farmers a more flexible selling regime and carry out pricing and marketing studies to develop the appropriate financial incentives/disincentives for livestock farmers. The Bank's attitude towards stricter enforcement of the Agricultural Resources Act (PCR pars. 9.07(b)) has been that if the land use mechanisms and financial signals to farmers are not conducive to sustained

P 1867 - 22 - use of the resources, little will be achieved solely through the regulating process.

4. Quantifiable benefits in the Second Livestock Project were to be derived from (a) incremental sales from ranches and (b) prevented weight loss from the cattle using an improved trek route system. The responsibility for monitoring benefits was to be handled by ILCA as part of its overall project contribution. However, while it had a continued in-country presence for the first two years of the project, staffing and funding difficulties at ILCA led to its Botswana office being closed down. No replacement monitoring input was provided for the project and consequently the PCR contains little or no information on benefits or losses. EAPSA staff feel that because project ranches did not become established in number until late in the project life, few benefits would at this stage be discernable. However, there is evidence that although four consecutive years of drought have reduced national livestock production, ranches have at least been able to maintain previous production levels. The droughts have had an equally serious effect on trek routes when grazing has ceased to exist, and farmers have stopped using this mechanism to move their cattle to the abattoir. When normal climatic seasons return livestock will exhibit some compensatory growth and as the project interventions are still technically valid, benefits can be expected to accrue.

5. As supplementary information we have included a Schedule of Estimated versus Actual Disbursements and the Basic Data Sheet17

1/ Estimated and Actual Disbursements attached and information on Basic Data Sheet included in the audit's Basic Data Sheet. - 23 -

BOTSWANA

Second Livestock Development Pt-ect (Ln. 1497-BT)

Project Completion levort

Schedule of Estimated vs ActutL Disbursements (US$ millior.

IBRD Fiscal Cumulative Itsbursements Actual as % Year and 2garter Estimated Actual of Estimate 1977 December 31 - -

1978 March 31 0.20 - June 30 0.40 -

1978 September 30 0.60 - December 31 0.80 - 1979 March 31 1.00 - June 30 1.25 0.03 2% September 30 t.50 0.08 5% December 31 1.75 0.11 8%

1980 March 31 2.00 0.25 13% June 30 2.25 0.37 16% September 30 2.50 0.71 28% December 31 3.00 0.78 26%

1981 March 31 3.50 0.98 28% June 30 4.00 1.06 27% September 30 4.50 1.17 26% December 31 5.00 1.25 25%

1982 March 31 5.50 1.38 25% June 30 6.00 1.48 25% September 30 6.20 1.59 26% December 31 6.40 2.05 32%

1983 March 31 6.45 2.94 46% June 30 6.50 4.31 66% September 30 - 4.38 67% December 31 - 4.43 68%

1984 March 31 - 5.03 77% June 30 - 5.41 83% September 30 - 5.77 89% December 31 - 5.81 89%

1985 March 31 - 6.02 1/ 93% 0 l/ US$ 0.48 million cancelled. 27 - 25 -

SECOND LIVESTOCK DEVELOPMENT PROJECT

PROJECT COMPLETION REPORT

GOVERNMENT OF BOTSWANA _ 才 (二 一

'州 · 、 州 以 方 才 牟 《抆 - 27 -

I* BACKGROUND

1.01 During the planning of the Tribal Grazing Land Policy, a major land reform, it was realized that there was a need for a project through which to implement the policy. Under the TGLP exclusive rights to tribal land were to be granted to groups and individuals In order to encourage the development of commercial ranches and at the same time to preserve the land rights of all. Only a few commercial ranches existed prior to the introduction of TGLP.

1.02 During the late 1960's the First Livestock Development Project was conceived under which commercial ranches were to be developed in the Western State lands. LPI was funded by IDA and $IDA and was Implemented during the planning of TGLP. The problems experienced during the implementation of LPI provided useful lessons for the design of the Second Livestock Development Project. The major difference between LP1 and LP2 Was that LPI ranches were developed on State land whereas the introduction of the TGLP allowed the development of ranches on tribal land that had been zoned commercial. In addition the ranches established by the Government on turnkey basis under LPI whereas ranches were developed bv the ranchers themselves under LP2. Individuals had begun to obtain de facto rights to grazing land through the drilling of boreholes which prevented the development of the communal areas for the majority of the population. The introduction of TGLP was intended to be a means of providing productive opportunities for all Botswana.

11. PROJECT FORMULATION

IDENTIFICATION AND PREPARATION

2.01 The proposal to implement a Second Livestock project was first discussed during an IBRD mission in February 1975. It was agreed that the proposal would be further formulated during the following First Livestock Development Project supervision mission. As a result of this mission a Project Preparation Team was established with the employment of the Team Leader being funded with LPI funds.

2.02 With the assistance of the IBRT) Regional mission in Eastern Africa, the Team Leader was appointed in November 1975. The former Animal Production Officer was seconded to the team on a full time basis while two water engineers assisted on a part time basis. The Project Preparation Team presented their final report in the form of .nApplication by the Government of Botswana for a Loan to Finance a Second Livestock Development Project in November 1976.

2.03 The project would be the major vehicle for the implementation of the Tribal Grazing Land Policy and as such would promote rational land use, commercial management and the efficient marketing of livestock. This would be complemented by management training and support to enhance the decision making and executive capacity of the livestock owners.

2.04 The components put forward by the preparation team were:

(a) the development of 150 commercial ranches and six pairs of communal grazing units; - 28 -

(b) the development and support of about 12 service companies that would provide management assistance and back-up services for member ranches;

(c) the provision of ranch development loans by the National Development Bank with possible assistance of the commercial banks

(d) the provision of loans by the National Development Bank for the purchase of immature fattening stock. These funds would be repayable on the sale of the cattle and the provision of further loans would not be automatic. This item would also provide a liquidity reserve for financing the development of ranches in the event of a serious drought;

(e) the improvement of the national livestock marketing infrastructure;

(f) the establishment of a fire control unit for protection of Government and communal properties, and to carry out extension on fire control matters;

(g) the establishment of a Ranch Management Training Center with two main functions: (i) to provide staff and facilities to train ranch managers, service company managers and ranch finance officers; (ii) to provide a ranch monitoring team which would also provide an assessment and recommendations concerning ongoing livestock development projects and assist in the preparation of further livestock projects;

(h) the provision of funds for minor communal projects, research and consultancies. 2.05 The total cost of the project was estimated at approximately P12 million, P6.2 million of which was to be funded by the IBRD. P702,2)U was to be funded by the European Development Fund and P2,186,628 to be funded by a private aid agency, Technoserve. The International Livestock Center for Africa had agreed to provide assistance in the monitoring and evaluation of the project. The remaining balance was to be funded by the Government.

APPRAISAL

2.06 The project was jointly appraised by IBRD, Technoserve, the European Development Fund and ILCA in January-February 1977. Two members of the team returned to Botswana in July, 1977 to complete the appraisal.

2.07 The Mission initially recommended that the project cover a five year period and include the following components:

(a).100 ranches to be developed on commercial land;

(b) development by the Animal Production Research Unit of 12 communal grazing units to test different grazing systems in-communal areas;

(c) provision of supervised credit for ranch development through the NDB and possibly the commercial banks; - 29 -

(d) provision of services and management support to ranchers through the formation of service companies and a strengthened Ministry of Agriculture extension service;

(e) development of livestock marketing infrastructure;

(f) establishment of a ranch training center;

(g) monitoring and evaluation of project progress and preparation of future projects.

2.08 The Appraisal Team introduced the following changes in the project design as put forward by the Preparation Tea-4:

(a) the number of ranches to be developed in the commercial areas was reduced 'rom 150 to 100. The revised estimate was based on the capacity of the institutions involved to demarcate and allocate land for ranch development;

(b) the idea of service companies was endorsed by the Appraisal Mission, however the proposal to establish a sizeAble number of companies each servicing 10-15 ranches was determined to be unworkable. An alternative proposal was put forward: the establishment of approximately 3 companies each servicing 30 ranches together with an improved extension service provided by the Ministry of Agriculture.

(c) the Preparation Team had envisaged that the Communal Grazing Units would be established where communities were particularly responsive and would be managed by the communities themselves through management committees. It was recommended that APRU provide management services to the communities while the CGU's would remain the property of the communities. In addition to the twelve Communal Grazing units a further two control cells were to be established;

(d) the First Control Unit was viewed as beyond the scope of LP2 and it was therefore recommended that the proposal be put forward as a separate project;

(e) the catt:, 'knance scheme was excluded from the project as it was believed that this would lead to further widening of the distribution of cattle ownership.

2.09 The Appraisal Mission was followed up with a letter from IBRD in May, 1977 in which certain conditions were laid down as prerequisites to loan negotiations. Firstly, since the implementation of LPI had implications for the success of LP2, IBRD requested GOB to undertake corrective actions to resolve the problems encountered during LP1 implementation. These included the reconstitution of the Botswana Livestock Development Corporation as a viable Government company, the auditing of the Livestock Project Management Unit accounts, the improvement of management support and extension services and the signing of the lease/loan agreements for the Nojane ranches. Secondly, IBRD attached considerable importance to the establishment of the Agricultural Credit Division within the National Development Bank as this would.be the - 30 - institution with the primary responsibility for project implementation. Thus IBRD requested GOB to undertake to establish the division within NDB, to recruit and train staff and to develop appropriate procedures for loan appraisal, approval and disbursement. In addition the relationship between the NDB and the commercial banks would need to be clearly defined with any on-lending agreements between GOB and the banks formalized in Subsidiary Loan Agreements. Thirdly, the introduction of the related legislation such as the Hypothecation Act and the Agricultural Management Associations Act would need to be well advanced prior to loan negotiations. The Hypothecation Act provided for the use of agricultural produce, including livestock, as security for loans. The Agricultural Management Associations Act provided for the constitution, registration and control of Agricultural Management Associations.

2.10 IBRD expressed concern over the possible negative effet on income distribution as the ranch development component was likely to benefit mainly the wealthier farmers. IBRD stressed the need to develop group ranches so that the smaller livestock owners would benefit from the project.

2.11 Prior to negotiations the following actions were undertaken: all the necessary legislation was passed, the BLDC was reconstituted as a separate government company, the au&Lt of the LPMU accounts was in progress and discussions had taken place with the commercial banks regarding their role in the implementation of LP2. The commercial banks expressed their willingness to participate in the project provided that the TGLP lease met their requirements, the lease proposed by the Ministry of Local Government and Lands would not allow the banks to use the ranch as security in case of default on the loan. The Land Board would retain the right to reallocate the land. At the time of loan negotiations the TGLP lease was still to be finalized by the Government. Regarding the role of service companies it was agreed that as Technoserve were no longer willing to participate in the project, the establishment of the Dong Dong Service Company on the Nojane ranches would serve as a test case to would determine whether service companies would be established under LP2. A Ranch Extension Unit was to be established in the Department of Agricultural Field Services in the Ministry of Agriculture to provide extension support to the ranch development program.

2.12 Loan negotiations took place in Washington from 19-23 September 1977 with IBRD Board approval granted and the Loan Authorization Bill passed through Parliament in December 1977.

PROJECT DESCRIPTION

2.13 The project was to form part of GOB's program for livestock development and would represent the first major investment in support of the TGLP. Although the Second Livestock Project differs in approach to the First and in the institutions involved, efforts were made to avoid the reoccurrence of the implementation problems encountered. During LP1 implementation officials became increasingly aware of the sociological and ecological issues involved in the implementation of a complex project. The principle means of addressing these issues was seen as the implementation of TGLP and LP2. Under LP2 the following components were to be financed: - 31 -

(a) Ranch Development. Individuals or groups of individuals having obtained leases to land in areas zoned commercial would be able to develop ranches through the provision of loans from the National Development Bank. A model for the development and management of a commercial ranch had been devised by APRU and would provide the basis for extension efforts and the assessmert of loan applications. Each ranch would cover approximately 6,400 hectares of land and would carry about 475 animal units. The main investments would include fencing, firebreaks, water developments, simple ranch buildings and working capital. The Ranch Extension Unit was to assist ranchers in drawing up detailed ranch plans.

(b) Communal Grazing Units. In recognition of the need to undertake research on communal area livestock management in experimental program was to be introduced to study the sociological and technical coustraints involved. Twelve grazing units were to be established under different ecological conditions to test a range of grazing systems. The exact management system to be employed had still to be formulated. The capital investments, such as fencing, firebreaks, water development and operating costs for the development period would be funded under the project. Proups of small cattle owners were to be encouraged to register as Agricultural Management Associations in order to become eligible to receive funding of the capital costs through the NDB. Operating costs would be covered through the payment of a P10 levy per member per annum. The establishment of AMA's would provide members with two benefits; registered groups would have a legal status and individual members would have limited liability.

(c) Credit for Communal Association Schemes. A line of credit would be made available to NDB to finance small scale projects proposed by Agricultural Management Associations. These would include schemes such as drift fences; improvement of watering facilities and marketing infrastructure.

(d) NDB Agricultural Credit Division. NDB would establish an agricultural credit division to strengthen and diversify its leading to the agricultural sector. The division would assist potential borrowers for ranch development loans in preparing ranch investment plans in conjunction with the Ranch Extension Unit of the Ministry of Agriculture, in completing loan applications and in providing a follow up service to ranchers. The main investment items would include construction of offices and housing as well as operating costs.

(e) Livestock Marketing Infrastructure. Three existing trek routes would be improved and 12 new trek routes developed, in order to assist farmers in marketing cattle. Physical improvements would include simple housing, kraals to rest and restrain cattle and improved watering facilities along the trek routes at intervals of roughly 20 km. Facilities at 15 railway loading stations would be improved to make transport of cattle more efficient and to reduce losses due to poor handling. Smallstock holding grounds would be developed at 5 sites in the Bobirwa area to assist cooperatives in smallstock marketing. (f

2.14 exchang rancher ranch m or the costs.

2.15 the pro to chan affect of farin limitat However the rat extent would b they wo provisi drought livesto substan changes because much as offset

2.16 expecte output the prc - 33 - were expected to benefit from the other components. The potential negative effect on income distribution was recognized as the project would most likely benefit the wealthier farmers and as no feasible solutions had yet emerged for communal livestock development. While the problems of communal livestock development were acknowledged, it was agreed that the development of a commercial livestock industry should not await their solution. The project's income distribution implications were to be carefully monitored throughout the life of the project.

III PROJECT IMPLEMENTATION

Effectiveness

3.01 Loan effectiveness was set for the end of March 1978 but the loan was not made effective until November 1978 due to delays in the finalization of the Subsidiary Loan Agreement and the Administrative Agreement to be signed between NDB and the commercial banks. Funds were made available from Government sources to enable certain components of the project to start without undue delay. The agreement with the EDF for financing the Ranch Management Center and the Evaluation Unit was signed on the 22nd June 1977. The agreement with ILCA to provide technical assistance in monitoring the project was signed on the 12th July 1977.

3.02 An implementation schedule was agreed upon by GOB and the IBRD appraisal mission and it was agreed that 1978/79 would be the first year of project implementation.

Progress with Ranch Development

3.03 Although the Agricultural Credit Division was fully staffed by 1979, the format for ranch plans, loan agreements and selection criteria could not be finalized and disbursement of loans initiated due to delays in the finalization of the TGLP lease and the Administration Agreement to be signed with the commercial banks. The final draft of the TGLP lease was approved by Cabinet in November 1978. However, delays were also experienced with the land use planning exercise to be undertaken by MLGL and the 4emarcation and allocation of TGLP ranches in commercial areas. These deLays were overcome to some extent with the recruitment of additional landuse planning staff. Problems with the use of the Sperry Navigator by the Division of Land Utilization for the demarcation. of commercial ranches were resolved. The first problem concerned the legal validity of the demarcation of land. This was solved by amending the Tribal Land Act. The second problem concerned the difficulty of cutting trace lines along distances of 8 kms in areas of thick bush. Practical solutions were provided to ranchers by the Division of Land Utilization. The process of allocation of ranches and approval of leases by the Minister of MLGL was also very slow; the allocation of 12 leases in Mid-Ngwaketse took 14 months. This process was facilitated by the publication of TGLP Guidelines for Land Boards which clearly laid out the procedures to be followed in the allocation of commercial ranches.

3.04 In 1981 it was agreed with the IBRD that because of the slow disbursement of funds, the ranch development allocation be reduced by P1,210,000. NDB undertook to widely publicise the availability of ranch development loans in conjunction with the Ranch Extension Unit. By 1983 many of the problems that had hampered the disbursement of loans were - 34 - resolved and there was a substantial increase in the demand for these loans. Due to the increased demand and the rising costs of ranch development from an estimated P33,000 to P40,000 in 1982/83 the allocation of P1 million was exhausted and an additional P1,035,000 was approved for this component. The major reason for the increased uptake was the extension efforts of NDB's Agricultural Credit Division and the Ranch Extension Unit which resulted in the target group beginning to appreciate the benefits of ranch development. By the end of the project period, a total of 78 loans had been approved over a 6 year period, 22 less than had been envisaged in the appraisal report over a 5 year period. There was little progress with the development of group ranches; by the end of the project period only 8 group ranches had been established compared to a total of 212 ranches owned by individuals and syndicates.

Progress with Communal Grazing Cells

3.05 The objective of this component was to test four grazing systems in three different ecological zones of communal areas. To assist in the dissemination of research findings to livestock owners in communal areas, to extend the activities of APRU to applied research and to prepare communities to assume responsibility for the management of the grazing cells.

3.06 The three ecological zones were to be studied were the hard veld, sand veld and the mopane veld. The communal grazing cells would consist of a peripheral fence around a water point, with all the management facilities located at the center. The encircling fence would prevent the use of the area by non-member cattle. More elaborate grazing systems were to be imposed upon this by the construction of further internal fences. The systems to be tested would involve: continues grazing, three paddocks per herd, nine paddocks per herd and eighteen paddocks per herd. Each CGC would be stocked with 7 to 12 month old weaners and after two years the castrates would be ready for slaughter and the females ready for service. Two control cells were incorporated into the component where all the grazing systems would be evaluated together in the same cell.

Communal Grazing Cells

3.06 The Appraisal report foresaw the establishment of 9 CGC's within the first year of implementation and the completion of the remaining 3 during the second year. During the first year it was only possible to establish a control cell at Makhi II and to identify communal groups interested in the scheme. Eight groups were identified and were in the process of registration as Agricultural Management Associations.

3.07 One of the reasons for the delay in the implementation of the component was lack of staff: the post of pasture agronomist was to be filled by the first year of project implementation. UNDP agreed to provide an agronomist under the UNDP/FAO Range Research Project, however the officer was only appointed in May 1980.

3.08 By the end of the project period only one Communal Grazing Cell had been established at Ntimbale although discussions had taken place with other communities. In particular communities at Rukuntsi, Lehututu and expressed interest in the establishment of CGC's in their areas and had registered as Agricultural Management Associations. However the - 35 - establishment of these three cells was thwarted by lack of water at each site. The CGC at Ntimbale became operational in 1981 and has proved to be successful in that the Cell has been stocked with 200 cattle owned by the 55 members and the benefits of controlled grazing have been demonstrated to the community. Member cattle received insemination service from the Ministry of Agriculture's Artificial Insemination Service and in 1984 40 mature steers were slaughtered at BMC, obtaining 5% super, 80% Grade 1, 13% Grade 2 and 2% Grade 3. The CGC has experienced water problems: the borehole reached a critically low level due to the drought. Funds were provided during 1983 to drill an alternative borehole and the community undertook to lay the pipes for the water reticulation. The success of the CGC at Ntimbale will be tested when the manager provided by APRU hands over the management responsibility to the members. They will need to demonstrate their long term commitment to the CGC through the provision of a manager and funds for the continuation of the project.

3.09 The major re&sons for the unsuccessful implementation of this component have been:

(a) Communities have proved unwilling to allocate land for the CGC's. It was difficult to arrive at a consensus on the siting of the cells. The conflict arose from the proposal to provide certain members of the community with exclusive rights to grazing in an area identified as overgrazed and by definition carrying an excess number of cattle. Communities would have been more receptive to the concept of CGC's if they had been given access to new grazing areas as in the case of Group ranches.

(b) Group formation proved to be a more difficult and lengthy process than originally envisaged. Groups had to register as Agricultural Management Associations as a prerequisite to funding being made available through the NDB. The formation of the AMA at Ntimbale covered a two year period from the point of identification to registration. APRU assisted the group Development Officers of the office of the Commissioner of AMA's in identifying and consulting with suitable communities. The concepts involved in group formation are new to the majority of communities in Botswana and the acceptance of these concepts and recognition of their benefits will therefore require more time than was envisaged in the Appraisal Report. During the implementation of this component it was realized that the target number of twelve CGC's was unrealistic and that a more flexible approach was required. In December 1983, due to the failure to establish additional CGC's and as it was unlikely that further Cells would be established prior to project completion. It was agreed to terminate the CGC component so that the unused funds could be reallocated to other components.

Control Cells

3.10 A control Cell was established at Makhi II in 1979 but was only half stocked with cattle purchased from the surrounding area due to the outbreak of Foot and Mouth Disease. Once fully stocked the Control Cell continued to operate satisfactorily and to provide research results on communal area livestock and range performance under different grazing systems. - 36 -

Progress with the Communal Credit Scheme

3.11 Agreement was reached in 1979/80 that funds would be provided to groups registered as AMA's to undertake livestock development projects on a 50% grant and 50% loan basis. However, suitable projects for this funding were not identified mainly because of the availability of funds from other sources on more favorable terms.

Progress with the establishment of the Agricultural Credit Division NDB

3.12 The establishment of the Agricultural Credit Division proceeded according to the implementation schedule. FAO agreed to provide a manager for the division who was appointed in September 1978. In 1978/79 the Regional offices at Maun, Francistown, Gaborone, Serowe and Kanye became operational. New premises were built initially at four of the Regional offices and the fifth Regional office was built in Gaborone in 1981. The growth of NDB's rural portfolio led to a decline in contact with borrowers and it was therefore decided to further decentralize the banks operations through the establishment of 12 credit offices. In addition, a decentralized loan approval scheme was introduced to minimize delays between loan applications and approvals. In 1980 the NDB General Manager was granted the authority to approve loans of up to P10,000 without prior Board approval and Regional Finance Officers were given authority to approve loans of up to P2,000. This limit of P2,000 was increased for Regional officers to P5,000 in 1981.

3.13 A training department was established in 1981 and a manpower training program for the bank drawn up. A training course was designed for Credit Officers to enable them to fulfill their tasks of assisting potential borrowers to make loan applications, assessment of loan applications and providing a follow up and an extension service to potential borrowers. By the end of 1983 62 officers had received in-house training. 14 officers received external training and 28 employees, mainly clerical staff, participated in academic and professional training. Peace Corps Volunteers were recruited to act as assistant Regional Finance Officers to allow time for training of Junior bank officers and to deal with the increased loan portfolio. At the end of the pro:1ect period the division was headed by a Credit Manager with degree training, with 5 trained Regional Managers, 12 District Credit Supervisors and support staff. Serowe and Kanye had employed Loan Recovery officers and plans were underway to provide these officers to the Selibe Phikwe and Ghanzi offices.

Progress with Trek Route Development

3.14 At the start of implementation of this component, the Department of Animal Health, the department responsible, was able to keep to the implementation schedule agreed with IBRD. The planned developments for the first year of project implementation necessitated an increase in the estimated expenditure from P96,000 to P216,000.

3.15 At the time of project preparation there was some controversy over the need to develop the "*thern Trek Route. It was agreed at that time that as the route was not heavily used and motorized transport was becoming more popular, the necessary repairs would not be included in LP2. However the outbreak of Foot and Mouth Disease changed the situation. It was - 37 - agreed that the route would need to be prepared for the increase in cattle movement once.the quarantine restrictions were lifted.

3.16 Increasing detailed district plans brought to light the limited space available for the development of trek routes. Thus it was decided that the MOA should formulate a National Trek Route Policy.

3.17 Prior to the approval of the National Trek Route Policy, the main priority was the improvement of the Northern Trek Route with work also undertaken on the Rakops-Makoba route. The Makoba-Palapye section was delayed while discussions were held on the siting of the Veterinary Cordon Fences. The routing of the eastern trek routes proved to be problematic as they were to be sited through heavkly populated arable areas. The National Trek routes Policy recommended the development of trek routes in the west of Botswana where iand pressure was not considered to be a problem. In eastern Botswana the lack of available grazing land was seen as a constraint. It was therefore recommended that holding and loading grounds were developed in difficult areas. Extensive negotiations were held with Land Use Planning Advisory Groups and Land Board on the siting of Trek Routes and to determine grazing rights between proposed watering points. These negotiations took two years and without such negotiations no meaningful implementation could have taken place. Due to the problem of land pressure in eastern Botswana the development of trek routes in this area was postponed.

3.18 The implementation of the trek route component gained momentum after the acceptance of the proposal to concentrate on the development of trek routes in the west of Botswana. Three routes were chosen: Shakawe-Maun, Ghanzi-Lobatse and Mabutsane-Khuis. The latter route was abandoned after drilling several blank boreholes and striking saline water in others. In the former two regions, 21 boreholes were drilled in each region and 9 and 10 equipped respectively. The Department of Water Affairs, the supervising agency, decided to equip boreholes in the Shakawe-Maun r,ute. The Ghanzi-Lobatse route was contracted to a private company as the DWA did not have the capacity to supervise all the regions. Kraals were constructed by the DAH at every completed watering facility to rest and restrain cattle en route. Other handling facilities provided include crushes, loading bays and watering troughs. Houses were also built by the DAN to provide accommodation to pumpers manning the boreholes.

3.19 Five vehicles were purchased during the project period and three were used by DWA for supervisory work. These vehicles were returned to the DAH to be used in trek route activities. Due to the failure of the drilling program in the Southern Kgalagadi equipping materials etc. intended for this section remained unused. All the engines and other scientific equipment remained with the DWA and other equipment will be used by the DA for maintenance purposes.

Progress with the Development of Railway Handling Facilities and Smallstock Holding Grounds

3.20 The Railway Handling Facilities component was completed in 1981/82 with the construction of 14 kraals, 9 of which were installed with water.

3.21 The implementation of the Smallstock Holding Ground component was delayed by problems of land allocation and low morale amongst cooperative - 38 - members due to the restrictions on smallatock sales to BMC during the FMD outbreak. Priority was therefore given to the Railway Randling Facilities component. Rolding grounds were finally completed at 5 sites during 1983/84 although the holding ground at Tonota was funded under the EDP funded project Services to Livestock Owners in Communal Areas.

3.22 All holding grounds were handed over to the Cooperatives, although the use of the facilities by the cooperatives has been less than envisaged. Some of the constraints that have been identified are:

(a) Although the management committees of the Cooperatives had accepted these facilities as essential for improvement of smal1stock marketing the cooperative members have not been made sufficiently aware of their benefits.

(b) Some of these societies had not previously marketed smallstock to the BMC and those that had, had been able to market their smal1stock without the use of holding grounds.

(c) The maintenance costs of the holding grounds were beyond the means of the societies. It was estimated that the society would require a throughout of 2,000 smallstock per year in order to meet maintenance costs.

3.33 It was therefore decided that the Division of Animal Production would maintain the facilities for a period that will allow the societies to establish themselves sufficiently to assume responsibility for the maintenance of the holding grounds. In the intervening period Smallstock Treatment Groups have been organized to use the handling facilities in the holding grounds with the objective of promoting the marketing of smallstock through the improvement in the quality of the animals.

Progress with the Provision of Technical Services

3.34 The technical services to be provided under the project consisted of a Ranch Management Center, monitoring and evaluation of the project progress and studies and experimental work.

Ranch Management Center

3.35 The Ranch Management Training Center was established at Ramatlabama to provide facilities to train managers of ranches, service companies, bank finance officers and extension staff involved in the implementation of TGLP.

3.36 The major building program commenced in October 1978 following changes in the design of the buildings. These design changes and rising costs necessitated additional funding: the estimated cost increased from P61,000 to P152,000. Repairs of the staff houses was completed in December 1978 and the offices were completed by February 1979 when the princ!.pal and his staff moved into them. Conversion of the old building into 8 offices, 2 classrooms, an open workshop, a large store and toilet facilities was completed by April 1979. The construction of the dormitory block and dining facilities was delayed by slow delivery of materials and lack of site supervision. This caused several postponements of the commencement of the first course. All 3f the construction was completed in June 1979. - 39 -

Demonstration facilities were constructed and procurement of vehicles and equipment completed the end of 1978. The final construction work was completed in 1981 with the provision of flooring, roofing and water tanks where these were required.

3.37 The principal of the center was in post from July 1978 and the three instructors, secretary and copy typist were in post from the end of December 1978. The bursar's post remained vacant until 1983 when the assistant bursar was promoted to the post of bursar. Initially the principal had to perform the bursar's duties which prevented him from fulfilling all his duties as principal.

3.38 The first course commenced in June 1979 with a total of 28 students. The names of potential candidates had been forwarded to the center by extension staff and places advertised in the press. The final selection was made by the Chief Animal Production Officer, the Research and Extension Coordinator and the Principal of the center. The center received the full intake of 30 students in 1981. Records have been kept of student performance in order to monitor changes in student performance. The sixth course commenced in February 1984 with an intake of 32 students indicating an increased demand for the courses provided at the center.

3.38 The Evaluation unit was established at the center in January 1979 with the staff supplied through a technical assistance contract for a period of approximately two and a half years. The employment of the team was financed by the EDF.

3.39 The Evaluation Team presented their work program in May 1977 when it was agreed that a number of sites should be selected to be studied in detail. The selection of the sites precluded any attempt at random sampling as the aim was to study different areas representing a range of characteristics including communal grazing areas, the TGLP zones, ecological zones and ethnic groups. Six areas of 35 km squarc each were selected at Maokane, Mosolotsane, Bobonong, Tsau, Samane FDA, Haina Veld FDA and a seventh site at Pelotshetlha was added to be studied in collaboration with the Integrated Farming Pilot Project and APRU. The total number of herds and cattle to be monitored amounted to 313 and 22,500 respectively.

3.40 The difficulties involved in the establishment of a system of monitoring livestock movement and performance were far greater than envisaged by the team. It was a year before consistent results were obtained, leaving approximately one year for the actual monitoring. The 1980/81 season could not have been considered a 'normal' year and the work undertaken by the team had to be viewed as a baseline against which to compare subsequent changes in the areas and herds surveyed.

3.41 The original Terms of Reference for the teti had been to "collect and collate basic data to monitor and evaluate the development emanating from the TGLP." Due to the delays in the implementation of TGLP which consisted mainly of planning and surveying during the period in which the team were in Botswana, the task of the team became redefined. It changed from one of evaluation to investigation of the basic features of communal area livestock management and provision of recommendations on appropriate methods of promoting the development of these areas. - 40 -

3.42 During project implementation two changes were agreed to: the inclusion of a sociologist and a larger budget for the employment of local enumerators and the extension of the date for submission of the Draft Final Report from let July 1981 to let September 1981. The final report was approved in October 1981 and recommendations were provided on, for example, resource management in communal areas, measures for improvement of the range and livestock, improving the understanding of TGLP in the CFDA's and reserved areas. The monitoring activities of the team were taken over by the Farm Management Unit within the Ministry of Agriculture.

Monitoring and Evaluation

3.43 The monitoring of the implementation of LP2 was the responsibility of an inter disciplinary team consisting of APRU, the Division of Animal Production, the Division of Land Utilization and the Rural Sociology Unit.

3.44 An agreement was signed with ILCA under which assistance was to be provided for the development of a program to monitor LP1 and LP2. An acting ILCA Country Representative was posted to Botswana in November 1977 and was replaced by a permanent Representative in June 1978.

3.45 The livestock economist provided by ILCA focused on three activities: implementation of the monitoring of LP2, assisting the Ministry of Agriculture to evolve a unified framework and methodology for monitoring livestock development projects, and monitoring the progress of the Nojane ranches developed under LP1. The representative also assisted the Range Ecology Unit in the development of methodologies for range monitoring and collection of livestock and economic data from the ranches by the REU. ILCA was involved in the development of a livestock simulation model which was applied in Botswana to analyze the economic trade offs between meat and milk production under different levels of supplementation.

3.46 Until April 1980 ILCA's work in Botswana was funded by ILCA, however due to a reduction in core funding GOB agreed to fund ILCA's activities in Botswana. GOB agreed to pay 80% of ILCA expenditure from LP2 funds until 31st March 1983. This period was later extended to 30th June 1983 whereafter ILCA's budget was funded by ILCA headquarters, apart from the aerial surveys which were funded from the investigative studies component.

3.47 In December 1980 the ILCA Representative was unable to return to Botswana and the post remained vacant for 6 months. During this period ILCA's monitoring activities either ceased or were taken over by other departments. From June 1981 to April 1982 an acting representative was in post and was replaced by the permanent representative in September 1983. The work program of the new representative included collaboration with two projects, the Integrated Farming Pilot Project and the Agricultural Technology Improvement Project, to undertake studies of grazing patterns and arable/livestock interactions.

3.48 Physical and financial progress of LP2 implementation was monitored by the Livestock Project Coordination Unit and the implementing agencies and annual reports were prepared by the Project Coordinator with the assistance of the ILCA Representative. In order to avoid duplication of M&E activities it was agreed that ILCA woul- assume responsibility for the monitoring of ranches developed in the commercial areas and that - 41 -

Evaluation Team at Ramatlabama would concentrate on developments in the communal areas.

3.49 A Senior Rural Sociologist was appointed in June 1980 with the task of monitoring the impact of TGLP. The RSU was involved in the nation wide consultation process which took place prior to the implementation of TGLP. The collection of socio-economic data from commercial areas provided the basis for the evaluation of the effects of TGLP implementation. With a major focus of this work relating to employees and residents of commercial ranches. The responsibility of the Rural Sociology Unit was to bring to the attention of policy makers these effects so that policy modifications and changes could be introduced. The Unit also worked in close cooperation with the agencies involved in the implementation of TGLP such as Land Boards, LUPAGS, the Land Division of MLGL and the National Institute for research. A list of studies and surveys conducted by the RSU is provided in Annex 3.

Investigative Studies

3.50 In 1981/82 it was agreed that funds would be provided to APRU to undertake stocking rate trials. Stocking rate recommendations were very limited for the wide variety of environments in Botswana. Trials were to be carried out at 5 APRU ranches to test 3 stocking rates: a light treatment (equivalent to understocking), a medium treatment (approximately the optimum stocking rate) and a heavy treatment (approximately communal area stocking rates). APRU would thus be in a position, in conjunction with the Range Ecology Unit, to assess the carrying capacity of land in different areas.

3.51 During 1982/83 the trials were reevaluated and subsequently modified. It was found that the range of the 3 stocking rates was too varied to produce useful results. The trials were altered so as to study more general trends such as the effects of the drought on range condition. In 1982 a Communal Grazing Simulation Trial was established at Morale to study the processes of range degradation in communal areas and to investigate the influence of 3 different levels of management upon livestock performance.

Preparation of Future Projects

3.52 Funds from LP2 were not utilized for the preparation of the successor to LP2, the National Land Management and Livestock Project.

IV THE IMPACT OF MANAGEMENT ON PROJECT ACHIEVEMENT

4.01 The management structure of LP2 was designed to fit into the existing Government structure. The implementation of each component was to be the responsibility of the department normally responsible for such activities. NDB was responsible for the disbursement of ranch development loans and funds for the Communal Grazing Cells. APRU was responsible for the establishment of the Communal Grazing Cells in conjunction with Group Development Officers and extension officers of the Department of Agricultural Field services and was involved in M&E activities. The Department of Animal Health was responsible for the development of Trek routes with the Department of Water Affairs responsible for the drilling and equipping of boreholes. The Division of Animal Production was - 42 - responsible for the development of Railway Handling Facilities and Smallstock Holding Grounds and the supervision of the Ranch Management Training Center. The Ranch Extension Unit provided extension se-.1ces to ranchers in commercial areas and the Division of Land Utilization was responsible for the demarcation of ranches.

4.02 The overall project policy formation, guidance and coordination was provided by two committees, the Livestock Development Steering Committee and a Livestock Project Coordinating Committee. A Livestock Project Coordinating Unit was established to serve as the secretariat for the two committees and a clearing house for project funds and accounts.

4.03 The Livestock Project Steering Committee was established in December 1976 before the IBRD appraisal. This high level committee comprised of three members: the Deputy Permanent Secretary MOA, the Director of Economic Affairs, Ministry of Finance and Development Planning and the General Manager NDB. The Project Coordinator acts as secretary to this committee to provide a feedback to the Coordinating committee of which he is the chairman. The Steering Committee was able to facilitate the progress of the project as its members had the authority to make decisions on issues such as the appointment of the Project Coordinator and the channelling of funds for the Communal Grazing Cells.

4.04 The Livestock Project Coordinating Committee was composed of those officers directly responsible for the implementation of the project. The main achievement of the committee was to provide information to all officials concerned with the implementation of the project and to highlight outstanding problems that required action. Where possible the committee made recommendations to the Steering Committee for resolving problems.

4.05 The Livestock Project Coordinating Unit was established in February 1977. Although the Project Coordinator was assigned to the post at that time his position was not confirmed until October 1978. Government decided to utilize existing staff in the Unit inspite of the need for a qualified accountant. The LPCU was responsible for following up on the recommendations of the two committees, assisting the implementing agencies in drawing up physical and financial plans, warranting of funds to the respective departments and the submission of the necessary documents to MFDP for reimbursement of expenditure by IBRD.

4.06 While delays in implementation of some of the components were outside, the control of either the LPCU or the implementing agencies, such as the delays in ranch allocations. The implementation of some components was impeded by the inability of the Project Coordinator to directly control the activities of the implementing agencies and to affect prioritization of activities within departments. The implementing agencies were responsible for the execution of other projects which in some cases were of higher priority and were therefore implemented at the expense of the implementation of LP2 components. For example, the Trek Route Officer was also responsible for the construction of Veterinary Cordon Fences which at the start of the project period were viewed by the DAH as a higher priority and thus work on the Trek Routes nas delayed. Likewise the officer responsible for the establishment of the Railway Handling Facilities and the Smallstock Holding Grounds was instructed to complete other developments prior to commencing work on these two LP2 components. The problem of lack of authority over implementing agencies was also clear when - 43 - the coordination of different departments was necessary as in the case of Trek Route development where the DAR, DWA and CTO were required to cooperate. The delays caused by this problem were resolved for the Trek Route component by the purchase of vehicles for the specific purpose of implementation of the component. However, the implementation of LP2 required less inter departmental cooperation than in the case of LP1. On the whole, the implementation of the components was effected fairly efficiently. The major constraint appears to have been a lack of suitably qualified staff.

Impact of Financial Factors on Project Implementation

4.08 The development costs of several components were in excess of the appraisal estimates due to both escalating prices, a problem made worse by delays in project implementation and by the exclusion of items from the component budget or revisions in the scope of activities. For example, the delays in the implementation of the ranch development component meant that at the time of development the costs had risen from an estimate of P22,020 at 1977 prices to P40,000 per ranch in 1982 and approximately P61,000 in 1984. Funds for the development of control cells under the Communal Grazing Cell component were not provided for in the appraisal estimate, it was agreed with IBRD to utilize funds from the tiallocated category. In 1980, a project revision was introduced whereby the unutilized funds for the Communal Credit Scheme were reallocated to the unallocated category for future utilization. In 1983 it was agreed to reallocate funds from the Communal Grazing Cell component to the Trek Route and NDB Agricultural Credit Division components as it was evident that further cells would not be established prior to project completion. Thus the slow disbursement of funds under several components allowed for the allocation of additional funds to components with a higher expenditure rate.

Procurements

4.09 The major constraints with regard to procurement were procurement of materials for the ranch development component. The ranchers, unlike under LPI, had the task of procurement of materials for ranch development. The problem of procurement had been foreseen at the time of project preparation and had been the rationale behind the proposal to establish service companies. The test case of the Dong Dong service company on the Nojane ranches had not proved to be a success thus the proposal to establish service companies on LP2 funded ranches was dropped. Procurement of materials for the other components was in accord with the IBRD requirements of competitive budding and no problems were experienced.

Costs -nd Disbursements

4.10 Actual project costs are given in Annex I, Table 1. The disbursement of funds was slower than envisaged in the appraisal report due to the delals in project implementation discussed in section. The closing date of the project was extended by one year to June 1984 to allow for the completion of the Trek Route component, to complete the establishment of three Communal Grazing Cells and to utilize the remaining funds to meet the increased demand for ranch development loans. - 44

Covenants

4.11 Most covenants of the Loan Agreement, BT 1497, were uncontroversial and were largely met by the Government. The two conditions that were not met during project implementation were 3.02 "the Borrower shall employ agricultural management and accounting consultants ani experts" and 3.07 "that NDB shall employ criteria for selection of borrowers that favor groups of livestock owners over individuals owning large herds in allocation sub-loans". As stated earlier an accountant was not provided to the LPCU and the condition of favoring groups of farmers over individuals proved to be unfeasible for NDB to implement.

V. AGRICULTURAL IMPACT

5.01 The project was designed to benefit livestock owners in several ways, the main objectives being to assist in developing commercial ranches in commercial areas, to seek new technologies to benefit small livestock owners, promote more efficient marketing of livestock and develop institutions to serve livestock development programs. The development of the livestock industry was to also promote the rational use of the range and thereby improve the condition of the range. The project was also seen as a vehicle for the implementation of TGLP, the objectives of this policy being:

To make grazing control, better range management and increased productivity possible. The improved management system must start with fenced areas and land over which exclusive rights are recognized. Therefore under certain conditions, groups and individuals must be granted exclusive rights to land. To safeguard the interests of those who own only a few cattle or none at all.

National Policy on Tribal Grazing Land 1975.

5.02 Thus the impact of the implementation of LP2 should be evaluated in the light of not only the objectives of the project itself but also with respect to the success of fulfilling the objectives of the TG,P which relate to increasing agricultural output on a sustained baste.

Ranch Development

5.03 The development of ranches in the areas zoned commercial was to lead to increased productivity on these commercial ranches through the introduction of improved management techniques and in the communal areas through a reduction in stocking rates as cattle were moved from heavily stocked communal areas to the commercial areas. Behind this reasoning lay two basic assumptions: that there were 'empty' areas in Botswana that would be zoned commercial and where commercial ranchers would move from the communal areas and that the granting of exclusive rights in the commercial areas would lead to the adoption of improved management techniques.

5.04 The TGLP zoning surveys in 1975-76 revealed that many of the areas of the country thought to be empty, which were to be zoned commercial already contained numerous water sources and substantial livestock and human populations. One of the problems with the discovery that large numbers of water sources existed in areas to be zoned commercial was that there would be little room for the large herd owners to move out of communal areas. Since boreholes existed in large numbers in - 45 - most areas zoned commercial particularly in Ngwaketse, Kweneng, Central and North West districts the process of establishing commercial areas has not resulted in a reduction of the stocking rates in the communal areas. In addition commercial ranchers are entitled to move their cattle back onto comm:nal land when grazing becomes scarce on the commercial ranches.

5.05 Improved herd and range management on commercial ranches has also proved to be quite limited to date, although the process of adoption of improved management is a long term one and ranch development therefore needs to be evaluated in the light of this fact. Some Land Boards attempted to introduce appendices to TGLP leases to establish stock limits on commercial ranches, however this was ruled to be illegal by the Attorney General's Chambers in 1979. The only requirement of TGLP ranchers is that they manage their ranches "in accordance with the principles of good husbandry".

5.06 By the end of the project period a total of 78 NDB loans had been approved and the development of the ranches had commenced. Although a data collection system was developed for the Ranch Extension Unit to monitor the establishment and operation of the ranches this system has not been implemented sufficiently to provide adequate data for a detailed evaluation. While the demand for ranch loans increased substantially during the last years of the project, the Ranch Extension Unit and the Rural Sociology Unit identified several problems and constraints faced by ranchers and extension staff.

(a) The development of ranches took place during a period of four years of drought which resulted in the decline in borehole yields on many of the ranches and caused degradation of the range condition of these ranches.

(b) Many of the ranches initially exceeded the recommended stocking rates of 400 LSU's per 6,400 hectare ranch. The overstocking of the ranches led to a serious decline in the condition of the ranches, a problem which was exacerbated by the drought. Many ranches for example in the Haina Veld remain overstocked. In areas where the condition of the ranches has deteriorated many ranchers have moved their cattle back to the communal grazing areas. Ranchers have retained the right to graze their livestock in the communal areas.

(c) The problem of degradation of range condition has been exacerbated by the construction of perimeter fencing at the expense of the estab'ishment of an efficient water reticulation system.

(d) Many of the ranchers have not employed qualified ranch managers to manage the ranches as the salaries required by the trained managers are prohibitively high. The ranchers feel that their debt burden prevents the payment of large salaries to ranch managers and ranches are therefore managed by untrained herdsmen employing traditional management system.s These herdsmen who have obtained considerable experience of livestock management are unable to attend the Ranch Management Centre as they do not possess the required academic qualifications. - 46 -

(e) The ranch managers that have been employed have resigned from their duties to seek higher salaries elsewhere.

(f) Very few ranchers reside on their ranches as they tend to be engaged in other income activities. This reduces the ability of the rancher to effectively oversee the management of the ranches. Ranch managers have complained that this situation is made worse by the fact that they are given little authority by the ranchers to take management decisions.

(g) Although APRU was able to provide the REU with a ranch development model few recommendations have been produced for appropriate grazing systems. The Ranch Extension Officers have received an education in general agriculture and are therefore not sufficiently trained in range management to advise ranchers on grazing systems.

5.07 Despite the constraints discussed above the impact of the provision of funds for ranch development cannot be assessed without taking into account the long term nature of this type of development and the severe impact of the drought.

Current Situation

5.08 The demand for the allocation of ranches has continued to increase the Land Boards have been put under pressure to allocate more ranches. As most of the previously unused land has been allocated Land boards are beginning to allocate ranches to existing cattle posts. This has led to increased conflicts between the potential ranchers and the other inhabitants of these areas.

5.09 The 1984/85 Annual Report of the Ranch Extension Unit reveals that as at the 1st February, 1985, a total of 83 ranches had constructed perimeter and internal fencing (see Appendix II).

Communal Grazing Cells

5.10 While the establishment of Communal Grazing Cells was unsuccessful in that only one communal cell and one control cell were established it was possible for the implementing agencies to draw conclusions from the implementation of this project. These observations and conclusions will assist in the development of future communal area livestock projects. The two cells also fulfilled valuable research and demonstration functions.

5.11 The conclusions arrived at by APRU were:

(a) solutions to the problems of communal area development cannot be imposed by Government from above but have to be arrived at by the communities themselves;

(b) the issues of stock control and grazing rights should only be raised when problems more easily perceived by the communities have been resolved; - 47 -

(c) boundary recognition is an essential prerequisite to voluntary stock control and is more easily instituted by cohesive groups; and

(d) development programs should work through local institutional frameworks where possible with realistic time scales.

Agricultural Credit Division - NDB

5.12 The provision of funds for the establishment of the Agricultural Credit Division within NDB resulted in a significant expansion of NDB's loan portfolio. NDB's lending to the agricultural sector increased from P2.8 million in 1979 to P19 million in 1983. NDB was able to increase its share of lending to the agricultural sector from 15.7% in 1979 to 55.9% in 1983.

Trek Route Development

5.13 The impact of the establishment of Trek routes is difficult to assess due to the effects of the imposition of livestock movement restrictions during the Foot and Mouth Disease from 1979/80 to 1981/82. The Dukwe and Makoba trek routes were not used during this period to trek cattle to the BMC abattoir in Lobatse however livestock were trekked along the Ghanzi trek route. The numbers of cattle arriving at BMC along the trek routes are given below.

Number of Cattle Trekked to BMC

GHANZI TREK DUKWE TREK MAKOBA TREK ROUTE ROUTE ROUTE

1980 4,278 1981 14,571 1982 27,748 - - 1983 19,188 34,720 9,458 1984 19,612 21,026 1,411

These figures do not reflect the numbers of cattle that are trekked to the line of rail and those cattle that are trekked for other reasons. The incidence of drought has led to an increase in the numbers of cattle slaughtered by BMC while at the same time the condition of livestock has deteriorated considerably. This may explain the initial increases in the number of cattle trekked and the subsequent decline in numbers as trekking resulted in losses as livestock condition worsened.

Railway Handling Facilities

5.14 Under this component existing facilities were upgraded at 15 sites along the line of rail. These facilities are thought to have improved the efficiency of livestock marketing and led to a reduction in losses during transportation. A total of 124,100 cattle were loaded at these facilities during 1984. - 48 -

Smallstock Holding Grounds

5.15 As discussed in paragraph 3.22 there was no direct impact from the establishment of Smallstock Holding Grounds under LP2, although this situation would change if these facilities were to be utilized in the future.

Ranch Management Training Centre

5.16 The Ranch Management Training Centre has provided training for ranch managers and extension staff. However as discussed in paragraph 5.06 ranchers have not employed ranch managers due to the high salaries demanded.

VI. TRIBAL GRAZING LAND POLICY

6.01 As discussed above the implementation of LP2 was directly affected by progress with the introduction of TGLP and therefore the validity of the rationale and assumptions behind TGLP also had an impact on success of LP2.

6.02 The concept of granting exclusive rights to tribal land was put forward in 1972 in a report on rural development undertaken by Dr. Chambers and D. Feldman. At the same time APRU produced research findings which indicated that it would be possible to substantially increase livestock productivity with improved livestock management techniques. The proposal to provide exclusive rights to tribal land was put forward in the 1975 White Paper entitled National Policy on Tribal Grazing Land. Land Boards were to divide tribal land into (a) commercial areas, where individuals and groups would be able to obtain leasehold rights for a period of fifty years, (b) communal areas, where the basis of land tenure would remain as before, and (c) reserved areas which would be set aside for future use by either stock holders and small livestock owners or wildlife, mining or cultivation.

6.03 During the implementation of TGLP the validity of the assumptions behind the policy was reassessed as discussed in paragraphs 5.04 and 5.05. This has led to changes in emphasis and stipulations of the policy. The major assumption concerning the availability of empty areas has proved to be invalid and has led to the dezoning of several ranches that had been zoned as commercial* Population surveys were conducted which revealed substantial populations residing in those areas thought to be empty. In addition one of the means of reducing overgrazing in the communal areas was the imposition of stock controls. However the consultation exercise that was undertaken prior to the implementation of the policy revealed widespread opposition to the proposal so that it was abandoned. The proposal to introduce stock controls in the lease was also dropped as discussed above. One of the major changes in the TGLP stipulations was rent for the commercial ranches was set at a sub-economic level; 4 thebe per hectare or P256 for a 6,400 hectare ranch. The intention had been to utilize the revenue from commercial ranches to develop the communal areas.

6.04 During 1980 there was an increased awareness of these issues which led to the commissioning of the report "Keeping an eye on TGLP" which provided to the Government advice and recommendations on the implementation of TGLP. The report recommended that the policy relating to the commercial - 49 - areas should remain unchanged but recommended that more emphasis should be placed on communal area development. The report further recommended the use of existing social institutions rather than attempting to overcome the perceived social constraints to communal area development. The recommendations of the report were accepted by the Government and the proposal to provide an increased emphasis was implemented by the development of Communal First Development Areas.

6.05 In 1983 the MLGL introduced additional guidelines for the implementation of TGLP. These related to issues such as compensation, allocation of boreholes, the enforcement of the Fencing Act, the development of communal service centers to provide for the needs of non-stock holders residing in the commercial areas, the location of ranches in relation to existing settlements and the introduction of a condition that ranchers effect ranch developments within the first three years of the lease period.

6.06 The Land Tenure Commission was appointed in May, 1983 with the task of 'reviewing the existing land tenure systems and advising Government on changes needed to improve them and remove any obstacles from the systems which were impeding development, whilst preserving and promoting the salient features of inheritability, security of tenure and access to land by all citizens.'

6.07 The Commission provided recommendations for all categories of land. The recommendations on grazing land were that no changes were required but that the solutions to the problems of overstocking and lack of communal area development lie with improved land use planning and changes in production methods and management practices. The Commission recommended that the Ministries of Local Government and Lands and Agriculture "intensify their efforts to develop workable solutions which can then be applied. Two additional recommendations were put forward regarding the restriction of movement of livestock from the commercial areas to the communal areas and the enforcement of the Conservation Act.

Sector Influences

6.08 The Second Livestock Development Project was implemented during a period in which there was an outbreak of Foot and Mouth Disease, a three year period of drought and escalating costs of certain materials, all of which had an effect on project performance. The Foot and Mouth Disease outbreak temporarily terminated the sales of beef to the European market and resulted in livestock movement restrictions both significantly affecting livestock revenues and beef exports. The Trek Route component was also affected by the disease outbreak as the construction of Veterinary Cordon Fences became a higher priority than the development of trek routes which were thus delayed. The drought affected mainly the viability of the commercial ranches and the Communal Grazing Cells, although the DWA had difficulty in finding high yielding boreholes for the trek routes.

VII. CHANGES IN REPEATER PROJECTS

7.01 In February 1983 a mission from the World Bank visited Botswana to assess the livestock sector and to identify projects suitable for Bank funding. It was agreed with the Government that an accelerated program for the implementation of TGLP was necessary for the development of the - 50 -

livestock sector and thus a project proposal was put forward and appraised by the Bank in September 1983.

7.02 In recognition of the need for a more effective land use planning as a prerequisite for the successful implementation of TGLP the National Land Management and Livestock project will strengthen the land use planning capability of the Government. Land Boards have not had sufficient technical expertise and therefore no comprehensive land use plans were produced. The major task of the TGLP implementation has been the zoning of land into commercial and communal areas. However as little basic data had been collected the actual land use was not always known at the time of zoning and some areas had to be dezoned. The project will address the problem of lack of trained land use planners through the upgrading of the Land Use Planning Advisory Groups to District Planning Units. This will be achieved by the employment of professional planners and a reduction of the large membership of the LUPAGS. Land use planning at the national and district levels will be coordinated through the employment of a Coordinator of Land Use Planning and the District Planning Units to be established in four districts: Central, Ngamiland, Kweneng and Southern, and Land Boards will be strengthened through the provision of housing, equipment and vehicles. The land use planning capabilities of the Ministry of Agriculture will be strengthened through the provision of funds for training of the Division of Land Utilization staff.

7.03 As discussed above the demand for ranch development loans increased substantially during the final years of LP2 implementation. Funds will continue to be provided as loans to ranchers through the NDB with which to develop commercial ranches. Groups of ranchers will be provided with grants for borehole drilling. In order to improve NDB's loan approval capacity two additional ranch finance officers will be recruited. Individuals will be encouraged to fund the drilling of boreholes from the sale of cattle.

7.04 One of the principal shortcomings of LP2 was the lack of progress with communal area development. The Communal Grazing Cell component met with little success as discussed above but did provide useful lessons for the design of future communal area projects. It became increasingly clear that greater community involvement was required in the planning of communal area livestock projects. A communal area livestock development group will be established within the Division of Animal Production, comprising a rural sociologist and two livestock deve-3pment specialists. Their task will be to identify and work with interested communities to evolve appropriate management systems.

7.05 The structure of the management of LP2 proved fairly efficient however there were several weaknesses in the coordination of project activities and the accounting mechanisms. To overcome these deficiencies a central Project Coordination Unit will be established as part of the Chief Agricultural Economist's office in MOA. The unit will be staffed with a coordinator, deputy coordinator, monitoring and evaluation officer, accountant and support staff. - 51 -

VIII. BANK PERFORMANCE

Identification

8.01 Identification of the Second Livestock Development Project formed part of the initial rlanning of the TGLP as LP2 was to be the principal means of implementing the policy. The formulation of the project proposal by the Bank and the Government was in accordance with the objectives of the TGLP.

Appraisal

8.02 The implementation targets for the components of LP2 proved to be over optimistic particularly with respect to the Ranch Development and Communal Grazing Cell components. The appraisal report envisaged the development 100 ranches over a two to three year period commencing in 1979. In practice there was little demand for ranch development loans until 1983 and at the end of the project period only 78 loans had been approved by the NDB. The appraisal report envisaged the development of 12 Communal Grazing Units: 9 in Year 1 and 3 in Year 2. During actual project implementation only one grazing cell and one control cell was established. The problems of group formation and acceptance of new concepts were not foreseen by the appraisal team. The costs of establishment of two control cells was excluded from the estimates of investment costs.

8.03 The assumptions made by the appraisal team related to the assumptions behind TGLP which did not always hold true, as discussed in paragraphs 5.04 and 5.05. The assumptions regarding improved productivity led to overambitious output and productivity coefficients when compared to the actual conditions on ranches developed under LP2.

8.04 There appears to have been little consultation with communities regarding the need for Smallstock Holding Grounds (see paragraph 3.22).

Major Covenants

8.05 The major covenants were largely in line with the actual implementation of the project apart from the condition regarding the provision of loans to groups of farmers in preference to individuals (see paragraph).

Supervision

8.06 There were a total of 14 supervision missions mounted by the World Bank, including the mission prior to loan negotiations in September 1977. This amounts to an average of two per year and includes two visits from the Bank Headquarters, one by the head of the Operations Unit and one by the head of the Operations Evaluation Unit in 1982.

8.07 In general the quality of supervision was good and major problems were clearly identified during review missions. - 52 -

IX. CONCLUSION

9.01 As discussed above the objectives of the TGLP were to be met through the implementation of LP2. TGLP can be seen as an attempt by the Government to reconcile the aims of increasing the national income through the commercialization of the livestock industry and the preservation of economic opportunities for all.

9.02 In order to promote the commercialization of the livestock industry increases in productivity were to be encouraged by the granting of exclusive rights to land. Groups and individuals would be encouraged to develop commercial ranches and to employ improved management practices. The establishment of commercial ranches would allow the movement of livestock from the heavily stocked communal areas. A communal area component was to be implemented under LP2 to demonstrate the benefits of improved management to communal livestock owners.

9.03 The Ranch Development and Communal Grazing Cell components of LP2 were to be the principal means of achieving the above objectives. The other components were to provide additional infrastructure and support services for the livestock industry.

9.04 The infrastructural and support components were successfully implemented and largely provided the benefits envisaged at appraisal with the exception of the Smallatock Holding Grounds and the Ranch Management Training Centre.

9.05 The implementation of the Ranch Development and Communal Grazing Cell components was less successful due to the more complex nature of the proposed developments. The success of the ranch development component relied on the assumption that the granting of exclusive rights to land would lead to the adoption of improved management systems. The rationale behind this assumption is that the communal system of land tenure does not encourage the individual to efficiently utilize the range resource. Only with the granting of exclusive rights is it within the livestock owners' interests to effectively manage the resource. This has not proved to be the case. Many ranches are still managed on traditional lines despite the construction of perimeter fencing which may be viewed by some ranchers only as a means of preventing the use of the land by other livestock owners. Many of the ranches have been seriously overstocked.

9.06 The validity of the assumption that exclusive rights will lead to improved management depends upon the fact that the resource is finite. In this situation it is in the ranchers' interests not to exhaust this resource. However ranchers have retained their access to the communal grazing areas so that when the exclusive resource has been depleted livestock can be removed. There is no incentive for the rancher to effectively manage a ranch under these conditions.

9.07 From the above discussion and the observations made during the implementation of LP2 it is possible to identify the requirements for the success of the ranch development component.

(a) The dual grazing rights of ranchers need to be modified as recommended by the Presidential Commission on Land Tenure. - 53 -

(b) Stricter control of stocking rates should oe enforced under the Agricultural Resources Act.

(c) The Ranch Extension Unit needs to be strengthened and Ranch Extension Officers provided with appropriate training.

(d) Further research is required to produce recommendations relating to grazing systems and other management practices.

9.08 The proposals for communal area development were not clearly defined under TGLP. Improvements in the communal areas were to be achieved through the movement of livestock from the communal areas. As discussed in section the development of commercial ranches has not led to the reduction of grazing pressure in the communal areas. The scope of the development under LP2 was limited and the implementation of the Communal Grazing Cell component met with little success. Conclusions have been drawn from the problems encountered with communal area development which will need to be taken into account in the design of future projects. - 54 - ANNEX 1

TABLE 1: COMPARISON OF APPRAISAL ESTIMATES AND ACTUAL EXPENDITURE

APPRAISAL ACTUAL COMPONENT ESTIMATE EXPENDITURE

Ranch development loans 2,210 2,210 Communal grazing units 1,227 331 Communal credit scheme 450 - NDB Agricultural Credit Division 1,243 2,905 Trek route development 566 2,604 Railway handling facilities and smallatock holding grounds 95 75 Ranch managing training center 591 254 Monitoring and evaluation 730 258 Project preparation 200 - Studies and investigative work 400 329 Contingencies 3,411 -

TOTAL PROJECT COST 11,124 8,772 - 55 -

ANNEX 2

TABLE 1: RANCH DEVELOPMENT PROGRESS AT February 1, 1985

RANCHES LEASES LOAN RANCH DISTRICT ALLOCATED SIGNED APPROVED DEVELOPMENT

MAUN 71 60 49 46PF/IF FRANCISTOWN 12 7 7 7PF/IF SEROWE 15 15 15 2PF KWENENG 17 15 14 12PF/IF NGWAKETSE 27 26 18 25PF/IF KGALASADI 24 20 2 2PF/IF GRANZI 27 27 27 27PF/IF

Total 193 170 132 121

NOTE: PF-Perimeter Fencingz IF-Internal Fencing

Table 2: GROUP RANCHES DEVELOPED UNDER LP2

NAME LOCATION MEMBERSHIP

Dutlhe 19 moshupa Mashamba 12 Itekeng Baruakgomo T1hobolwane 17 Selehalo Sekoma 10 Bolkago Southern District 20 Tshipidi Southern District 11 Longwe Nata 10 Muzwibabe Zwenshambe Nata 10 -56 - ANNEX 3 Page 1

RURAL SOCIOLOGY UNIT TGLP REPORT SERIES

TITLE AUTHOR DATE

TGLP Baseline survey M.J. Odell June (TGLP Consultation campaign, results of the A. Etherington 1977 baseline survey)

A study of attitudes to Agricultural Problems, Y. Merafe Jan the TGLP and Group Formation Part 1, the North 1977 West District

A study of attitudes to Agricultural Problems, B. Tsimako Dec the TGLP and Group Formation Part IV, Central 1979 District

A study of attitudes to Agricultural Problems, P. Ntseane June the TGLP and Group Formation Part V, Southern 1980 District

A study of attitudes to Agricultural Problems, P.G. Oganne 1977 the TGLP and Group Formation Kweneng West

Seminar Report: Socio-Economic monitoring of M.J. Odell Sept TGLP 1977

Nolane Village and the Nolane Farms, Ghanzi P. Kiaer-Olsen Jan District: Discussion Papers G. Khumalo 1980 S. Bekure A. Kgosidintsi

A Preliminary Report on some Socio-Economic 0. Gulbrandsen Jan aspects of land use in the communal areas of 1980 the Bangwaketse

TGLP Group Formation: Report of Senior M.J. Odell Nov Agricultural Officers Workshop Y. Merafe 1976

A Preliminary Report on Group Formation 0. Gulbrandsen April within the Ngwaketse South Extension Area U. Bergen 1977 RSU

Draft Report to Government of Botswana on G. Haaland Oct socio-economic monitoring of the TGLP 1977 -57 - ANNEX 3 Page 2

TITLE AUTHOR DATE

Report on the Socio-Economic aspects of G. Khumalo March ranch development Nolane Farms 1978

Report of a survey conducted in Ngamiland Y. Merafe July on the socio-economic impact of Foot and 1978 Mouth Disease

A Report of a study conducted in Ngamiland Y. Merafe Sept on the socio-economic impact of the P50 1978 Advance Loan Scheme and Flood Relief Scheme - 58 - ANNEX 4

SCHEDULE OF ESTIMATED VS ACTUAL DISBURSEMENTS (US$ '000)

IBRD Fiscal Cumulative Disbursements Actual as % Year and quarter Estimated Actual of Estimate

1977 December 31 - - 1978 March 31 200 - June 30 400 -

1978 September 30 600 - December 31 800 - 1979 March 31 1,000 - June 30 1,250 30 2%

1979 September 30 1,500 81 5% December 31 1,750 114 8% 1980 March 31 2,000 254 13% June 30 2,250 366 16%

1980 September 30 2,500 713 28% December 31 3,000 781 26% 1981 March 31 3,500 985 28% June 30 4,000 1,065 27%

1981 September 30 4,500 1,172 26% December 31 5,00') 1,251 25% 1982 March 31 5,500 1,384 25% June 30 6,000 1,486 25%

1982 September 30 6,200 1,393 22% December 31 6,400 2,056 32% 1983 March 31 6,450 2,942 46% June 30 6,500 4,317 66%

1983 September 30 - 4,289 66% December 31 - 4,437 68% 1984 March 31 - 5,034 77% June 30 - 5,419 83% - 59 - ANNEX 5

SELECTED REFERENCE DOCUMENTS

1. Second Livestock Development Project Progress Reports 1978/79-1983/84, Ministry of Agriculture.

2. An Application by the Government of Botswana For a Loan to Finance a Second Livestock Development Project, November 1976. Livestock Project Preparation Team, Ministry of Agriculture.

3. Government Paper No. 2. National Policy on Tribal Grazing Land, July 1975.

4. Keeping an Eye on TGLP July 1980. Stephen Sandford, National Institute of Research.

5. An Evaluation of Livestock Management and Production in Botswana, January 1982, Carl Bro International.

6. An Approach to Group Ranching and the Management of Communal Grazing in Botswana. R.J. Sweet and B.L. Addy APRU.

7. Report of the Presidential Commission on Land Tenure, December 1983.

8. Annual Report of the Ranch Extension Unit (Draft) 1984/85 Ministry of Agriculture. v - 0 40 0 h-

121N 方 F〔SRUARY1986 化 IBRD 18776 u- Kmungula Z A M 8 1 A

sk Ch BOTSWANA 0?8p NATIONAL LAND MANAGEMENT C 08E AND LIVESTOCK PROJECT

Ä7< Planning Area* Under the New Project

Dovolopment Areas Under Sacond Livestock pmjact First Livettock Z I M B A B W E rc,ýgmentt Arsug Under

f .4.ýr810x1-m4,LI>APÅI8K National Park*

Ex Isting Trek Routes VE R ~A Vsterinary Cordon Fonces Bulowayo vaccin* instituto

1~ 1 PA ut Sub-Depot$

Livestock Advisory Centers

swo gwebana Export Atatt*irt c 0 Proposed Abettolf tSI 0

-POW PrȊsod r stown Or000 etih no A&

4ý ä" ma fe CENTRAL n

~RJF ÅG ont Drift

#T

U,

stidsflong S erwood RanCh

10 c

EXISITINGROADS- RITUMINOUSSURFAr£D MAINROAOS GRAVELSURFACEO NAIN ROAD$ NAINfEEDER ROAD$ 1 SECONDARYftOAOS Y0 ------DISTRICTROAX Lefitiakong L 10 MA N ROADS UNDER CONSTRUCTION TO 181TUMINOLISPAVED STANDARD 24- GA i RAILWAYS RIVERS DISTRICT BOUNDARIES I tie Mochudilj INTERNATIONAL 130UNDARItS oaggdit b011 be Sikwane

0 30 onye cir n~# I E. tse ANGOLA ZAMBIA 101ý130418 ------.4o ýZIMBABWE

Romet NAMIBIA BOTSWANýý- RANCH rRA Mafeking CENZER 0 5~ 261 SWA Z l tA ID_ 0 10 '00 SOUTH AFRICA KILOMETERS, 0 SOUTH u. AFRICA

216« 2,8* MAY 1985