List of Tables, Figures and Maps

Final Report
Socio-Economic Household Survey: Integrated Community-Based
Eco-System Management
Prepared by
Social Impact Assessment and Policy Analysis Corporation
(Pty) Ltd. (SIAPAC)
for the
Ministry of Environment and Tourism
October, 2006

Table of Contents

List of Tables, Figures and Maps

1Introduction

1.1Introduction

1.2Aim and Objectives

1.3Structure of the Report

2Methodology and Approach

2.1Overview

2.2Approach

2.3Methodology: Start-Up and Training

2.3.1Inception

2.3.2Start-Up and Design

2.3.3Sampling

2.3.4Training

2.3.5Pre-Testing

2.3.6Final Arrangements Prior to Fieldwork

2.4Methodology: Implementation

2.4.1Field Team Management

2.4.2Field Team Procedures

2.4.3Detailed Field Report – Kunene Region

2.4.4Detailed Field Report – Caprivi Region

2.5Methodology: Data Entry/Validation, Analysis and Write-Up

2.5.1Data Entry/Validation

2.5.2Data Analysis/Report Write-Up

3Key Findings

3.1Introduction

3.2Household Assets

3.3Livestock, Small Ruminants and Poultry Ownership and Production

3.4Crop Production

3.5Restrictions on Resource Use

3.6Participation in Conservancy

Annex A: Terms of Reference

Terms of Reference

Consultant’s Understanding of the TOR

Annex B: Field Instruments

List of Tables, Figures and Maps

Tables

Table 21: Replaced PSUs in the Epupa Control Area

Table 22: Questionnaires Completed in Epupa Control Area

Table 23: Questionnaires Completed in Ehirovapuka Conservancy

Table 24: Questionnaire Completed in Purros Conservancy

Table 25: Questionnaires Completed in #Khoadi //Hoas Conservancy

Table 26: Questionnaires Completed in Torra Conservancy

Table 27: Questionnaires Completed in Mayuni Conservancy

Table 28: Questionnaires Completed in Kwandu Conservancy

Table 29: Questionnaires Completed in Salambala Conservancy

Table 210: Questionnaires Completed in Kasika Conservancy

Table 211: Questionnaire Completed in Kabulabula Control Area

Table 31: Household Assets (ownership)

Table 32: Livestock, Small Ruminant and Poultry Resources

Table 33: Crops Produced

Table 34: Plans and Activities

Figures

Figure 31: Restrictions on Resource Use

Figure 32: Knowledge of Local Conservancy

Figure 33: Household Member Involvement in Conservancy

Maps

Map 1: Map of Namibia Highlighting Kunene Region and Caprivi Region

Map 2: Map Showing Conservancies in Kunene Region

Map 3: Map Showing Conservancies in Caprivi Region

- 1-

1Introduction

1.1Introduction

In 2006 a socio-economic household survey was conducted for the Integrated Community-Based Eco-System Management (ICEMA) Project. The ICEMA Project is funded by the Global Environment Facility, through the World Bank, as well as the French Global Environmental Fund, and is implemented by the Ministry of Environment and Tourism (MET). The Project aim is to support “community-based integrated ecosystem management practices through the National [Community-Based Natural Resource Management] CBNRM framework … to … restore, secure and enhance key ecosystem processes in targeted conservancies, with biodiversity and land conservation and sustainable use as a goal” (Terms of Reference).

As part of the monitoring and evaluation requirements of the ICEMA Project, one key measure is to establish the varied impacts conservancies have had on rural livelihoods and local economic activity. Terms of Reference (TOR) for a survey were therefore issued in May, 2006, and awarded in June to Social Impact Assessment and Policy Analysis Corporation (SIAPAC), a Namibian socio-economic research firm. SIAPAC reported to a Reference Group, while an advisor from the Environment Department at the World Bank (Dr. Sushenjit Bandjopadhyay) designed the two questionnaires, subsequently revised and finalised by MET, the World Bank, and SIAPAC. There were some changes to the Terms of Reference, reflected in the Final Inception Report. The TOR, and the comments on the TOR, are noted in Annex A.

The survey was conducted in Kunene Region in the far north west, and Caprivi Region in the far north east. These regions are indicated in the following map:

Map 1: Map of Namibia Highlighting Kunene Region and Caprivi Region

Sampled conservancies are indicated in the following map:

Map 2: Map Showing Conservancies in Kunene Region

Map 3: Map Showing Conservancies in Caprivi Region

1.2Aim and Objectives

The aim of the study is to provide an assessment of the impact of community conservancies on the poverty and welfare of rural households’. The objectives of the study were as follows:

“extend and improve the database on household economies generated by the previous survey, by providing more detailed economic information on households in conservancy areas and by including households from non-conservancy areas for comparison;
facilitate a deeper understanding of the factors that determine how individuals and households spend their time and money through the analysis of income, expenditure, consumption and time use data, collected from a wide range of different households;
provide a basis for quantitative comparisons between the livelihood strategies of households in the areas where conservancies have been established, and the strategies of households outside those areas;
examine whether households’ livelihood strategies are changing over time, as conservancies become more established;
illuminate policy issues concerning the rural household economy (labour market policies, wages and income policies, equity) and how these issues are affected by access or lack of access to important natural resources”.

1.3Structure of the Report

The report serves as a record of the conduct of the study, and also includes analysis across key variables, presented by conservancy; the bulk of the analysis will be conducted by the World Bank. Therefore, the structure of the report is as follows:

Chapter 1Introduction and Overview

Chapter 2Methodology and Approach

Chapter 3Key Findings

Field instruments are included in Annex B.

2Methodology and Approach

2.1Overview

In this section, the approach for the field investigation is described, and methods used to obtain relevant information presented.

2.2Approach

Primary data collection involved the use of two approaches:

Administration of a highly-structured quantitative questionnaire to 960 heads of households;
Administration of a total of 66 community level quantitative interviews to village leaders and Conservancy Management Committee members.

2.3Methodology: Start-Up and Training

2.3.1Inception

An Inception Report was submitted as the first deliverable for the consultancy. The purpose of the Inception Report was to detail the agreed upon approach for project implementation. The Inception Report was finalised after the following meetings:

The Inception Meeting between the Project Management Committee and SIAPAC was held on 13 June 2006 in the MET Conference Room. The meeting covered the following main agenda items: 1) background to the project; 2) implementation plan; 3) development of data collection instruments; and 4) involvement of the World Bank Data Analyst.
A Questionnaire Development Meeting was held on 27 June 2006 in the MET Meeting Room. The meeting was attended by 3 representatives of SIAPAC, 3 representatives of MET and 1 representative of the World Bank. The main purpose of this meeting was to discuss the implementation of the household questionnaire and community-based questionnaire based on results from the pre-test.
Continuous discussion took place with regards to project training and implementation between SIAPAC and MET.

The above meetings contributed towards the development of the Draft Inception Report and the Final Inception Report, the latter submitted to MET in final form in July, 2006.

2.3.2Start-Up and Design

An Inception Meeting was held on 13 June 2006 in the MET Conference Room. At this meeting, the following were considered:

Background to the Project – Dr. Barnes from MET gave background information on the reasons for the field investigation. He noted that the WILD Project had planned a socio-economic impact assessment of conservancies from the year 2000. With assistance from a consultant from Sweden, a questionnaire was prepared. However, because of varied interests and high expectations from a variety of sources, the socio-economic impact assessment issues of interest to the economists in MET were severe diluted. In the end, there were many aspects of the survey, implemented in 2002, that were not measured in sufficient detail, and in a sufficiently useful fashion, to meet the information needs of the economists in MET. Nevertheless, the 2002 and the 2006 surveys did cover the same areas, although the 2002 survey did not include a control group. The ICEMA Project offered an opportunity to collect the information required, as part of its monitoring and applied research component. This survey represents one of the key data collection activities under IDEMA. The information to be collected by the survey is specified by MET, and is meant to meet the specific needs of MET.
Implementation Plan – Mr. Mouton from SIAPAC outlined activities by dates. He highlighted the particular importance of securing census data for the survey locations in a timely manner, given problems that had arisen in the past in this regard. He noted that the schedule as shown in the Proposal was still in place, and described activities underway to meet these deadlines.
Involvement of the World Bank Data Analyst – MET noted that a primary reason for the survey was that the data will ultimately be analysed econometrically. The Client will be assisted in this by the World Bank’s Environment Department. Thus, a Data Analyst from the World Bank would arrive on 17 June in Namibia to assist with the survey.
Data Collection Instrument Development – A draft Quantitative Questionnaire was provided to the Consultants on 12 June, and discussed at the Inception Meeting in broad terms. It was noted that the Consultant should take the draft Quantitative Questionnaire and prepare a pre-training version of the questionnaire by 19 June. Discussions were also held about the Focus Group Discussion (FGD) Instrument. It was noted that the main purpose of the FGD Instrument was to provide information that would contextualise information provided through the Quantitative Questionnaire. A related purpose was to collect ‘conservancy level’ data that would yield important data on prices and standardise units of measurement.

While the Consultants developed a draft household questionnaire and a draft focus group discussion instrument, with the arrival of the World Bank Data Analyst, both of these were altered significantly. In the end, the focus group discussion instrument was replaced by a community level leader questionnaire (included in Annex A).

A Questionnaire Development Meeting was held on 27 June 2006 in the MET Meeting Room. In this meeting, the following were discussed:

The main purpose of this meeting was to discuss the results of the pre-test that took place on 26 June 2006 at the Tsiseb Conservancy in the Erongo North and Kunene South regional areas. This pre-test covered the household questionnaire developed by the World Bank Data Analyst. The findings of the pre-test were: 1) the administrative length on the household questionnaire was at least 1.5 hours, and in some cases over two hours; 2) some indicators such as quantifying use of resources were problematic; 3) cancelling the planned key informant interview instrument and the inclusion of the community-level quantitative questionnaire.
The meeting agreed that the household questionnaire was too long in administrative length and that it needed to be reduced by 30% to be implementable. The meeting studied the questionnaire and reduced it by approximately 30%.
The meeting agreed that more time needed to be allocated to training to ensure that all trainees understood, in the same manner, indicators dealing with the quantification of resources used. The meeting agreed that the community-level quantitative questionnaire would be administered to 68 Village Headmen and 8 Conservancy Management Committee representatives.

2.3.3Sampling

An equal probability sample was drawn within each strata in the selected conservancy areas, while all PSUs within the control areas were automatically included. Across strata, over- and under-sampling took place, allowing for strata-level analysis, with data weighting to accommodate over- and under-sampling to be carried out by the World Bank Data Analyst.

It was originally planned to have a quantitative sample size of 1,100. This sample size was reduced to 960 quantitative interviews in order to provide for the following:

an additional 68 community-level quantitative interviews;
a longer household quantitative questionnaire in administrative time; and
the complexity of the household quantitative questionnaire.

The above sample size was discussed and agreed upon by the World Bank Data Analyst and MET.

The issue of securing 2001 census data and census maps was highlighted at the Inception Meeting. With these maps and the corresponding data, organised by PSUs, a proper sample was be pulled. Data obtained from the Central Bureau of Statistics were:

total number of households for each Primary Sampling Unit (PSU) in the sample locations;
conservancy maps with PSUs across each conservancy; and
maps showing the ‘match’ between the sample locations and the PSU.

From available 2001 census data, a Probability Proportionate to Size (PPS) sample was drawn within the conservancy area, while all PSUs in the control areas were automatically included. Effectively, this meant that each unit of analysis (in this case, a household) had an equal opportunity to appear in the sample (within each of the PSU listings). This meant that some larger PSUs had multiple clusters selected for the sample, and that thirty clusters per zone were randomly identified using the PPS procedure across 73 PSUs (26 PSUs in the Kunene Area and 47 PSUs in the Caprivi Area).

Within each cluster, households for interviews were randomly selected. Because of the lack of village locations and names within PSUs, each PSUs boundary was first established on the ground. The data collection team, with conservancy management members and village leaders, therefore established PSU boundaries. The centre of each selected PSU was identified from where ‘spinning the bottle’ random selection procedure was use to identify a random starting direction. Each enumerator selected a random number between 1 and 10 to determine the interval for the first house to be selected for interview. If a household was abandoned, or otherwise unavailable for interview (in the case of temporary absences, up to three repeat visits were made to try and secure an interview), these households were substituted with the household to the right of the closest entry/exit point to the household where the interview was conducted.

A few sampling challenges were created by the lack of key data and circumstances on the ground.

It proved impossible to collect data on all locations within each PSU, because this information was unavailable within the CBS census database. This was especially the case in conservancy areas in the eastern parts of the Caprivi Region (the floodplain area). In some PSUs, only one or two major villages were indicated.
Floods in the eastern Caprivi floodplain area made it extremely difficult to reach PSUs in Kasika and Kabulabula Conservancies.

2.3.4Training

Training was originally planned to involve training the field officers to administer the household questionnaire and KIIs. Because of changes in the types of data collection tools, this section is divided into two: 1) training household questionnaire and 2) training community level questionnaire. Training took place from 22 to 29 June 2006, including Saturday and Sunday.

2.3.4.1Training for the Household Questionnaire

A total of 2 Field Survey Co-ordinators, 4 Field Supervisors and 16 Enumerators were employed for fieldwork (four teams of five members comprising 1 Field Supervisor and 4 Enumerators, overseen by a Field Survey Co-ordinator who oversaw two teams). Language skill requirements were as follows:

Two teams of Otjiherero- and Damara-speakers (both also speaking Afrikaans) for Kunene and Erongo regions.
Two teams of siLozi-speakers for Caprivi Region.

To ensure that a sufficient number of experienced and trained field officers were available for fieldwork, the Consultants identified field officers from its extensive database of officers it had employed on previous surveys. For the 16 enumerators needed for fieldwork, a total of 18 were identified to attend training, from which the best 16 were selected.

Training took place at the NamPower Convention Centre in Windhoek, while pre-testing took place in Tsiseb Conservancy that runs across northern Erongo and southern Kunene regions. Training lasted nine days, including the pre-tests. This meant that the Damara and Otjiherero pre-tests were conducted, but that siLozi pre-testing could only be done among trainees (with non pre-test in Caprivi Region). Training lasted one day longer than originally planned because of the complexity of the field instruments.

Training was conducted by the World Bank Data Analyst, SIAPAC’s Team Leader and SIAPAC’s Field Survey Co-ordinators. Mr. Mouton, SIAPAC’s Team Leader, attended training virtually full time due to questionnaire complexity. The two other SIAPAC trainers thereafter served as Field Survey Co-ordinators for the northeast and northwest. The World Bank Data Analyst, Dr. Sushenjit Bandyopadhyay, attended training full time. Training included the following:

An overview of the investigation – Mr. Mouton gave an overview of the investigation, while Dr. Jon Barnes provided background information to CBNRM, the ICEMA programme and goals and objectives of the household survey.
A discussion of the basic principles of quantitative and qualitative research – detailed discussion were held regarding the basic principles of research, conducted by Mr. Mouton.
Training on field sampling procedures – this was a straightforward task as all trainees had implemented this type of sampling strategy before.
A detailed review of the Household Questionnaire. Survey Co-ordinators and trainees participated actively in training. Each question was discussed in detail until all trainees understood the purpose behind the questions, potential responses and the manner in which questions should be asked. An extended set of discussions took place for questions that were difficult in structure and nature, such as the following:

Need for name of household head and GPS coordinates of household locations and issues regarding confidentiality.
Structure of the questionnaire prohibited control by supervisors in terms of marking codes.
Questionnaire was considered too long for implementation.
Structure of the questionnaire was problematic for implementation.
Measurement of distances to facilities.
Reliability of expenditure responses.
Measuring quantity of products.
Measuring crop production and income from it.
Measuring use of natural resource as income.

Collective solutions were found to concerns raised with regard to the above. The World Bank Data Analyst was satisfied with the way forward in terms of measuring the above-mentioned indicators. The Consultants still had some important concerns, in particular regarding the accuracy of overall measures of income and expenditure. Despite this, every effort was be made to ensure that the results of the survey were acceptable.