Baseline Report for the Impact Evaluation of the Central ...

62
Baseline Report for the Impact Evaluation of the Central African Republic Support to Vulnerable Groups Community Development Project (P111679) June 2012 Radu Ban (The Bill & Melinda Gates Foundation) Matthias Rieger (The Graduate Institute of International and Development Studies, Geneva) Felipe Alexander Dunsch (The World Bank) Marcus Holmlund (The World Bank)

Transcript of Baseline Report for the Impact Evaluation of the Central ...

Page 1: Baseline Report for the Impact Evaluation of the Central ...

Baseline Report for the Impact

Evaluation of the Central African

Republic Support to Vulnerable Groups

Community Development Project

(P111679) June 2012

Radu Ban (The Bill & Melinda Gates Foundation)

Matthias Rieger (The Graduate Institute of International and Development Studies, Geneva)

Felipe Alexander Dunsch (The World Bank)

Marcus Holmlund (The World Bank)

Page 2: Baseline Report for the Impact Evaluation of the Central ...

2 PDCAGV IE Baseline Survey Report

Acknowledgements We would like to acknowledge especially the following individuals who have made important

contributions to the work presented here: Evariste Simbarakiye, PDCAGV Project Coordinator; from the

Central African Institute of Statistics and Economic and Social Studies (ICASSES), Mustapha Issen (former

Director General), Serge Mathinide (current Director General), Jonas Nangola (Technical Director of the

Baseline Survey), M. Perkyss (2nd Technical Director of the Study), and Jean-Barthelemy Mbaitar (Data

Entry Supervisor); and from the World Bank, Bernard Harborne (Task Team Leader), Paul Bance

(Operations Officer), Christopher Saunders (Operations Analyst), David Tchuinou (Senior Country

Economist), and Carine-Reine Mbdeo Ngassia (Receptionist).

Page 3: Baseline Report for the Impact Evaluation of the Central ...

3 PDCAGV IE Baseline Survey Report

Abbreviations

ADI Africa Development Indicators

CAR Central African Republic

CDC Commune Development Committee

CDD Community-Driven Development

ICASEES Central African Institute of Statistics and Economic and Social Studies

IE Impact Evaluation

LDF Local Development Fund

PDCAGV Support to Vulnerable Groups Community Development Project

PRF Priority Response Fund

VDC Village Development Committee

Page 4: Baseline Report for the Impact Evaluation of the Central ...

4 PDCAGV IE Baseline Survey Report

Contents Acknowledgements ....................................................................................................................................... 2

Abbreviations ................................................................................................................................................ 3

Résumé ......................................................................................................................................................... 8

Executive Summary ..................................................................................................................................... 11

Introduction ................................................................................................................................................ 14

1. Impact Evaluation of the Central African Republic Support to Vulnerable Groups Community

Development Project .................................................................................................................................. 15

1.1 PDCAGV ....................................................................................................................................... 15

1.2 Research questions ..................................................................................................................... 16

1.3 Impact evaluation design .................................................................................................................. 17

Table 1.3.1: Phasing of project components by evaluation group ..................................................... 17

1.4 Data collection .................................................................................................................................. 20

Table 1.4.1: Target sample by survey type ......................................................................................... 20

Table 1.4.2: Target and final sample by survey type .......................................................................... 21

2. Baseline survey coverage .................................................................................................................... 21

2.1 Geographic coverage .................................................................................................................. 21

Figure 2.1.1: Geographic Coverage ..................................................................................................... 21

Table 2.1.1: Geographic Coverage ...................................................................................................... 22

3. Household and individual-level results ................................................................................................... 22

3.1 Household size, age, and gender distribution .................................................................................. 22

Table 3.1.1: Household demographics ............................................................................................... 22

Figure 3.1.1: Population (in millions) .................................................................................................. 23

Figure 3.1.2: Poverty Rates ..................................................................................................................... 23

3.2 Principal activities of household members (overall and by gender) ................................................ 25

Figure 3.2.1: Principal household-member activities ......................................................................... 25

Figure 3.2.2: Agricultural dependency ............................................................................................... 26

3.3 Illnesses preventing work/going to school ....................................................................................... 27

Table 3.3.1: Reasons for absence from work ...................................................................................... 27

3.4 Education status of household members ......................................................................................... 27

Table 3.4.1: Education indicators ........................................................................................................ 28

Figure 3.4.1: Net primary school enrollment ...................................................................................... 28

Page 5: Baseline Report for the Impact Evaluation of the Central ...

5 PDCAGV IE Baseline Survey Report

3.5 Health/nutrition indicators ............................................................................................................... 29

Table 3.5.1: Selected health indicators ............................................................................................... 29

Figure 3.5.1: Under-5 mortality rate ................................................................................................... 30

3.6 Type of dwelling ................................................................................................................................ 30

Figure 3.6.1: Type of dwelling ............................................................................................................. 30

Figure 3.6.2: Roof material ................................................................................................................. 31

Figure 3.6.3: Wall material .................................................................................................................. 31

3.7 Time to access key infrastructure ..................................................................................................... 32

Table 3.7.1: Time to access locations of interest (average by foot in minutes) ................................. 32

3.8 Ethnicity/religion ............................................................................................................................... 33

Figure 3.8.1: Ethnicity ......................................................................................................................... 33

Figure 3.8.2: Mother tongue ............................................................................................................... 34

Figure 3.8.3: Religious affiliation ......................................................................................................... 34

3.9 Perception of village problems ..................................................................................................... 35

Figure 3.9.1: Village problems ............................................................................................................ 35

3.10 Prevalence of theft, violence, and conflict ..................................................................................... 36

Table 3.10.1: Prevalence of violence and theft .................................................................................. 36

3.11 Personal security ............................................................................................................................. 37

Table 3.11.1: Personal security perceptions ....................................................................................... 37

3.12 Participation in village meetings/volunteerism .............................................................................. 37

Table 3.12.1: Participation in village affairs ........................................................................................ 37

3.13 Trust ................................................................................................................................................ 38

Table 3.13.1: Who would you ask for help in case your mobile phone gets stolen? ......................... 38

3.14 Attitudes.......................................................................................................................................... 38

Table 3.14.1: Personal attitudes ......................................................................................................... 39

3.15 Nganga contacts/supernatural powers .......................................................................................... 40

Table 3.15.1: Nganga contacts/supernatural powers ......................................................................... 40

4. Village-level results ................................................................................................................................. 41

4.1 Village size ......................................................................................................................................... 41

4.2 Village access .................................................................................................................................... 41

Table 4.1.1: Condition of main village access road ............................................................................. 41

Figure 4.1.1: Village access ................................................................................................................. 41

Page 6: Baseline Report for the Impact Evaluation of the Central ...

6 PDCAGV IE Baseline Survey Report

4.3 Pygmy populations around villages .................................................................................................. 42

Figure 4.2.1: Pygmy population .......................................................................................................... 42

4.4 Availability and condition of key infrastructure in villages ............................................................... 43

Figure 4.4.1: Key infrastructure .......................................................................................................... 43

Figure 4.4.2: Infrastructure rehabilitation and construction needs ................................................... 45

Table 4.4.1: Primary types of infrastructure used in neighboring villages ......................................... 46

5. Commune-level results ....................................................................................................................... 47

5.1 Commune leadership ........................................................................................................................ 47

5.2 Commune transparency .................................................................................................................... 47

Table 5.2.1: Commune councils and the public .................................................................................. 47

5.3 Commune Budget ............................................................................................................................. 48

Figure 5.3.1: Commune budget .......................................................................................................... 48

5.4 Commune tax revenues .................................................................................................................... 48

Figure 5.4.1: Tax collection and sharing with prefectures ...................................................................... 49

5.5 Principal agricultural products of communes ................................................................................... 50

Table 5.5.2: Principal agricultural product of communes ................................................................... 50

Source: Commune survey ........................................................................................................................ 50

5.6 Community organizations ................................................................................................................. 51

Figure 5.6.1: Community organizations .............................................................................................. 51

5.7 National/international NGO presence in communes ....................................................................... 52

Table 5.7.1: Presence of national NGOs, international development programs, local development

plans .................................................................................................................................................... 52

6. Village Chief Survey ................................................................................................................................. 53

6.1 Basic summary statistics ................................................................................................................... 53

Table 6.1.1: Village Chief Summary statistics ..................................................................................... 53

Table 6.2.1: Average age and years in office ...................................................................................... 53

6.2 Ascent to power ................................................................................................................................ 54

Figure 6.2.1: Assumption of power ..................................................................................................... 54

6.3 Chief roles ......................................................................................................................................... 55

Figure 6.3.1: Chiefs’ participation in community groups .................................................................... 55

Table 6.3.1: Other administrative positions of village chiefs .............................................................. 55

Table 6.3.2: Party membership of village chiefs ................................................................................. 56

Page 7: Baseline Report for the Impact Evaluation of the Central ...

7 PDCAGV IE Baseline Survey Report

Figure 6.3.2: Village chief membership in commune council ............................................................. 57

6.4 Village conflict ................................................................................................................................... 57

Figure 6.4.1: Village conflict ................................................................................................................ 57

7. Commune Mayor Survey .................................................................................................................... 58

7.1 Demographics ................................................................................................................................... 58

Table 7.1.1: Commune mayor demographics ..................................................................................... 58

Figure 7.1.1: Commune mayor and household ethnicity ................................................................... 58

Figure 7.1.2: Commune mayor and household religion ..................................................................... 59

7.2 Education .......................................................................................................................................... 60

Figure 7.2.1: Education of commune mayors ..................................................................................... 60

Figure 7.2.2: Comparison of commune mayor and household head education levels ...................... 60

Table 7.2.1: Literacy of commune mayors .......................................................................................... 61

7.3 Commune council membership and other leadership functions ..................................................... 61

Figure 7.3.1: Mayors’ role in commune council ................................................................................. 61

Figure 7.3.2: Mayors’ assumption of office in commune council ....................................................... 61

Table 7.3.1: Traditional chieftaincy and length of tenure .................................................................. 62

Table 7.3.2: Political party membership and family involvement in public administration ............... 62

Page 8: Baseline Report for the Impact Evaluation of the Central ...

8 PDCAGV IE Baseline Survey Report

Résumé Ce rapport décrit les résultats de l’enquête de base de l’évaluation d’impact du Projet de

Développement Communautaire et d’Appui Aux Groupes Vulnérables (PDCAGV), financée par

l’Association Internationale de Développement (IDA) pour la République Centrafricaine. Ce projet vise la

réhabilitation de l’infrastructure sociale et l’amélioration de la capacité des parties prenantes locales

dans la planification et gestion communautaire du relèvement local.

L’enquête de base a été exécutée par l’Institut Centrafricain des Statistiques et des Etudes Economiques

et Sociales (ICASEES) avec l’appui technique du Development Impact Evaluation Initiative (DIME) de la

Banque Mondiale. Ce rapport présente les principaux résultats de cette enquête.

L’enquête a été administrée en Octobre 2010 dans 64 communes et324 villages, et à 324 chefs de

village, 3 201 ménages, et 64 maires de commune, dans les dix préfectures du sud du pays. L’enquête

de base comprend cinq modules: ménage, village, commune, chef de village et maire de commune.

Questionnaire des ménages

- 52.8% des membres de ménages gagnent leur vie en travaillant dans l’agriculture ou l’élevage.

- 73% vivent au-dessous du seuil de pauvreté de XAF 318 (ca. USD 0.65) par jour.

- 31% des adultes sont lettrés (47% des hommes ; 16% femmes). Ce chiffre est au-dessous de la

moyenne nationale (58% en 2008).

- 61% des enfants entre 0 et 5 ans ont un retard de croissance et 64% sont mal-nourris.

- L’Accès à l’eau potable est rapporté comme la première priorité pour le développement local

(31%), suivi par l’amélioration des infrastructures et services de santé (23%), l’éducation (17%)

et les routes (10%).

- 26% des membres des ménages n’ont pas pu travailler ou se rendre à l’école pour cause de

maladie durant le mois précédent l’enquête. Parmi eux, 28% n’ont pas eu accès aux services de

santé et 25% sont allés dans un hôpital qui, en moyenne, était à 93 minutes à pied. Les plaintes

les plus communes sont la diarrhée (21%), le paludisme (13%) et la fièvre (13%).

- 59% des ménages ont assisté à une réunion de village durant les 12 mois précédents l’enquête.

- 85% des ménages ont déclaré qu’ils sont disposés à travailler volontairement pour la

communauté.

- 25% des ménages ont été victime de vol de bétail et 18% ont été victime de vol des

possessions personnelles durant les six mois précédents l’enquête.

- 29% des répondants (incluant hommes et femmes) ont dit que le mari avait le droit de frapper

sa femme quand elle « se comporte mal ».

- 87% des répondants affirment que les femmes peuvent être des bons politiciens et qu’il faut les

encourager de se présenter comme candidats aux élections.

- 99% des répondants désirent des élections démocratiques et libres, mais seulement 42% disent

que la violence n'est jamais justifiée pour poursuivre des buts politiques.

- 6% des répondants ont un membre de leur famille qui a été blessé ou tué a cause de conflit

durant les trois années précédentes l’enquête.

Page 9: Baseline Report for the Impact Evaluation of the Central ...

9 PDCAGV IE Baseline Survey Report

- 17% des répondants disent qu’il est normal de payer un pot-de-vin à un fonctionnaire du

gouvernement pour l’encourager.

- Le chef de village reste l’autorité la plus crédible. 68% des répondants disent qu’ils

s’adresseraient à lui en cas de vol de téléphone portable; la deuxième réponse la plus fréquente

étant « Je ne demande pas d’aide. »

Questionnaire des Villages

- L’accès aux villages du projet est un grave problème. Des 324 villages inclus dans l’enquête, 28%

sont accessibles seulement par véhicule 4x4 et avec de grandes difficultés. 10% ne sont

accessibles qu’à pied. Seulement 27% des villages ont accès au réseau de téléphonie cellulaire.

En moyenne, il faut marcher 98 minutes pour accéder à une zone de couverture du réseau

téléphonique.

- L’infrastructure dans les villages du projet est gravement limitée. Plus que 50% des villages

n’ont pas accès a une source d’eau potable (en moyenne, il faut marcher 30 minutes pour

accéder à une source) et moins de 2% ont accès aux toilettes améliorés. Près de 85% des

villages n’ont pas accès à un hôpital et plus que 60% n’ont pas d’école (en moyenne, il faut

marcher 44 minutes pour accéder à une école). Il faut marcher 162 minutes pour arriver à un

marché et 109 minutes pour accéder au réseau électrique.

- Les pygmées vivent dans le voisinage de 9% des villages visés par le projet, mais ils sont peu

présents dans ces mêmes villages.

- 18% des villages contiennent des personnes déplacées à cause de conflit.

- 16% des villages avaient des incidents de violence entre habitants du village durant les six mois

précédents l’enquête. 11% de villages ont subi des épisodes de violence causée par des

personnes extérieures du village. En plus, 7% des villages ont eu une personne blessée ou tuée à

cause de conflit durant les trois années précédentes l’enquête.

Questionnaire des Communes

- Des 64 communes inclues dans l’enquête, 66% révèlent leur budget au publique et 76% publient

les notes de leur réunions.

- 41% de communes ont rapporté un budget zéro/négligeable/non applicable en 2009. Ce

nombre tombe à 22% en 2010.

- Environ 45% des communes ont un plan de développement local ou des ONG actives.

- Les sources principales des revenues d’impôts sont les licences pour vendre au marché et les

actes d’enregistrement civil. Les revenues d’impôts sont partagés avec les préfectures dans 40%

des cas.

Questionnaire des Chefs du Village

- 94% de chefs du village sont des hommes, et 45% ont un père qui était lui-même chef. La

moyenne d’âge d’un chef de village est de 51 années, et en moyenne la durée de son mandat

est de 9,5 ans.

- 61% des chefs de village rapportent qu’ils ont été élus.

Page 10: Baseline Report for the Impact Evaluation of the Central ...

10 PDCAGV IE Baseline Survey Report

- Selon les chefs, l’affiliation aux partis politiques est faible. 96% disent qu’ils ne sont membre

d’aucun parti politique et 76% rapportent qu’ils n’ont pas de membres de leur famille dans une

officine publique.

- L’adultère (28%), le vol des possessions et de bétail (26%), les disputes sur la distribution de

l’eau (25%) et les dettes (25%) sont les causes principales de conflit entre les villageois.

Questionnaire des Maires de Commune

- La composition ethnique des maires de commune correspond avec la composition ethnique

des ménages enquêtés. Concernant les affiliations religieuses, les minorités sont

surreprésentées parmi les maires: seulement 76,6% des maires sont chrétiens, mais le

pourcentage correspondant pour les ménages est de 87,2% ;. 14% des maires de commune

(représentant 6.4% des ménages) sont musulmans; et 9,4% des maires appartiennent à d’autres

religions. L’âge moyen des maires de commune est de 50,1 ans.

- La grande majorité des maires (81%) ne sont pas titulaires d’un autre poste.

- Le taux d’alphabétisation des maires est élevé. 79% savent lire et sont capables d’écrire une

lettre en français. Les membres d’un ménage moyen, en revanche, ont un taux d’alphabétisation

de seulement 31% (dont 16% pour les femmes).

- Les maires de commune ont aussi un niveau plus élevé d’éducation formelle par rapport aux

chefs de ménages. 86% des maires ont terminé leurs études primaires. Le chiffre correspondant

pour les chefs de ménage est seulement de 32%.

Page 11: Baseline Report for the Impact Evaluation of the Central ...

11 PDCAGV IE Baseline Survey Report

Executive Summary This report describes the baseline survey results for the impact evaluation (IE) of the IDA-funded Central

African Republic (CAR) Support to Vulnerable Groups Community Development Project (PDCAGV, from

Projet de Développement Communautaire et d’Appui Aux Groupes Vulnérables; P111679), which aims to

rehabilitate social infrastructure and improve the capacity of local stakeholders to plan and manage

local recovery in targeted areas of CAR.

The baseline survey was carried out by the Central African Institute of Statistics and Economic and Social

Studies (ICASEES, from Institut Centrafricain des Statistiques et des Etudes Economiques et Sociales),

with technical assistance from the Bank’s Development Impact Evaluation Initiative (DIME). This

baseline report highlights key results from the baseline survey.

The survey was carried out in October 2010 and included a total of 324 villages, 324 village chiefs, 3,201

households, 64 communes, and 64 commune mayors in the ten southern préfectures covered by the

project. The baseline survey comprised five modules: household, village, commune, village chief, and

commune mayor. Key results from each module are highlighted here.

Household Survey

52.8% of surveyed household members make a living from agriculture and/or animal husbandry.

73% of surveyed household members live beneath the national poverty line, with average daily

per capita income of XAF 318 (about USD 0.65).

31% of adult survey respondents are literate (47% of males; 16% of females). This is lower than the

official national average, reported as 55% in 2008.

61% of infants between the ages of 0 and 5 are stunted and 64% are malnourished.

Access to safe water is reported as the top development priority (31%), followed by the

improvement of health infrastructure/services (23%), education (17%), and roads (10%).

26% of household members were unable to work or go to school due to illness in the month

preceding the survey. Of these, 28% did not use health care services. 25% of respondents sought

care at a health post which, on average, took 93 minutes to reach (on foot). The most common

complaints were diarrhea (21%), malaria (13%), and fever (13%).

59% of households attended a village meeting during the 12 months preceding the survey.

85% of households claim to be willing to volunteer for the benefit of their community.

25% of households had been victims of theft of animals and 18% had been victims of theft of

personal goods during the 6 months preceding the survey.

29% of survey respondents (including females and males) think that beating one’s wife is

legitimate if she is “not behaving well.”

87% of respondents answer yes to the question, “Women can be good politicians and need to be

supported to take political posts.” But there are very few cases of a woman as village chief.

99% of respondents expressed a desire for free democratic elections, but only 42% answered “yes”

to the question, “Violence is never justified to achieve political goals.”

Page 12: Baseline Report for the Impact Evaluation of the Central ...

12 PDCAGV IE Baseline Survey Report

6% of respondents have had a member of their family killed or injured due to conflict in the past

three years.

17% of respondents believe “it is normal to bribe officials to help speed up the work.”

The village chief remains the most trusted authority, with 68% of respondents stating that they

would call him/her for help were they to be robbed. The next most common response was, “I

wouldn’t ask for help.”

Village Survey

Access to project villages is a major concern. Of the 324 villages covered by the survey, 28% could

only be reached by 4x4 vehicle with great difficulty, while 10% could only be reached by foot. Only

27% of villages have mobile phone reception (average travel time to access mobile network: 98

minutes on foot).

Infrastructure is severely limited. Over 50% of villages do not have access to improved water

sources (average travel time to nearest source: 30 minutes) and less than 2% have access to

improved toilet facilities. Nearly 85% of villages do not have a health facility and more than 60% do

not have a school (average travel time to nearest primary school: 44 minutes). It takes an average

of 162 minutes to reach a market, and 109 minutes to access electricity.

Pigmies live in the vicinity of 9% of target villages but are often less present in the villages

themselves.

18% of villages contain conflict-displaced people.

16% of villages had cases of violence among village inhabitants in the past 6 months. 11% of

villages saw violence committed by persons not from the village. Also, 7% of villages had a person

injured or killed due to conflict in the past 3 years.

Commune Survey

Of the 64 communes for covered by the survey, 66% make their budget publicly available and 76%

publish the minutes of meetings.

41% of communes had a zero/negligible/not available budget in 2009, falling to 22% in 2010.

About 45% of communes have a local development plan or active NGO operations.

The principle sources of tax revenue are market rights and civil registry, and tax revenues are

shared with the prefecture in around 40% of cases.

Village Chief Survey

94% of village chiefs are male, and 45% of respondents’ fathers were also village chiefs. The

average village chief is 51 years old and has held his post for an average of 9 years and 6 months

years.

61% of village chiefs report that they were elected to office.

According to the chiefs, political party membership is low. 96% say that they are not affiliated to

any political party and 76% state not having any relatives in public office.

Adultery (28%), theft of goods and livestock (26%), disputes over water (25%), and debt (25%) are

the most common causes of conflict among villagers.

Page 13: Baseline Report for the Impact Evaluation of the Central ...

13 PDCAGV IE Baseline Survey Report

Commune Mayor Survey

The ethnic composition of commune mayors matches the ethnic composition of surveyed

households. In terms of religion, however, minorities are disproportionately represented among

mayors: 87.2% of the surveyed households, but only 76.6% of mayors, are Christian. 14.1% of

council mayors are Muslim (compared to 6.4% for households) and 9.4% follow other religions.

Average age of commune mayors is 50.1.

The vast majority of mayors (81%) do not to hold any other public position.

Literacy levels among commune mayors are generally high. 79% can read and draft a letter in

French. This is in contrast to the average household, which has a literacy rate of 31% (for women,

this falls to 16%).

Commune mayors have higher levels of formal education than do household heads. 86% of mayors

had completed at least primary education; the corresponding figure for household heads is 32%.

Page 14: Baseline Report for the Impact Evaluation of the Central ...

14 PDCAGV IE Baseline Survey Report

Introduction This report describes the baseline survey results for the impact evaluation (IE) of the IDA-funded Central

African Republic (CAR) Support to Vulnerable Groups Community Development Project (PDCAGV, from

Projet de Développement Communautaire et d’Appui Aux Groupes Vulnérables; P111679), which aims to

rehabilitate social infrastructure and improve the capacity of local stakeholders to plan and manage

local recovery in targeted areas of CAR. This survey was carried out in September and October 2010, in

328 villages across 64 communes in 10 prefectures in the south and south-west of the country. The

survey was implemented by the Central African Institute of Statistics and Economic and Social Studies

(ICASEES, from Institut Centrafricain des Statistiques et des Etudes Economiques et Sociales). This report

describes the main findings.

As the PDCAGV is initially covering parts of the country only, a rigorous impact evaluation is being

carried out to assess its potential for scale-up, its suitability as a mechanism to target development

funds towards vulnerable populations, its community-level impacts with respect to local governance

capacity and improvement of the living situation of vulnerable groups, and its performance relative to

other forms of local service delivery.

The PDCAGV consists of three operational components: (a) capacity strengthening for local

development, (b) a priority response fund (PRF), and (c) a local development fund (LDF). The IE uses an

experimental design to explore the causal impact of these components on household, village, and

commune-level outcomes, and was designed to provide methodologically rigorous answers to three sets

of questions at different stages of program implementation. First, it was to analyze the impact of

capacity strengthening (component A) on the institutional capacity and perceived legitimacy of

commune governments, immediately after the first year of the project and before any service delivery

projects had been carried out. Second, it was to compare the implementation performance of centrally

prescribed development projects under the PRF to the implementation performance of projects with

local ownership established through a participatory decision-making process under the LDF. Third, it was

to evaluate the overall impact of the program at the level of households, villages, and communes.

Following the progression of project implementation, the IE will focus initially on the rehabilitation of

water pumps under the PRF (component B) to examine the social and economic impacts of

infrastructure rehabilitation under this component. As such, the focus of this baseline report is on

household, village, and commune level outcomes which may be important to assessing the impact of

the PRF. That said, these and other results from the baseline survey remain relevant for the possible

evaluation also of capacity strengthening for local development (component A) and the LDF (component

C), and the IE design remains in place to answer all three research questions, contingent on

developments in project implementation.

This report is divided into six sections. Section 1 briefly describes the PDCAGV and its IE. Section 2

describes the geographical coverage of the baseline survey. Sections 3 through 7 present key results

from the household, village, commune, village chief, and commune mayor surveys, respectively.

Page 15: Baseline Report for the Impact Evaluation of the Central ...

15 PDCAGV IE Baseline Survey Report

1. Impact Evaluation of the Central African Republic Support to

Vulnerable Groups Community Development Project

1.1 PDCAGV The project development objective of the PDCAGV project is to rehabilitate social infrastructure and

improve the capacity of local stakeholders to plan and manage local recovery in targeted areas of CAR. It

is expected that the project will improve the overall social and economic situation of the targeted

communities, particularly rural communities, and that through their direct participation in the decision-

making process and implementation of sub-projects, communities will regain some level of trust in local

and central authorities, hence contributing to improved governance. In so doing, the project will

contribute to peace-building and particularly to rebuilding the social contract between citizens and their

government.

The project is financed by a US$8.0 million International Development Association grant with a further

US$12 million provided by the African Development Bank (US$ 20 million total). It consists of four

components.

1. Component A: Capacity Strengthening for Local Development. The capacity building component

is designed empower rural communes and villages to prepare and implement development

plans in an inclusive manner, with adequate support from Government staff. Activities under

the component address the technical and fiduciary skills needed at the different decentralized

levels to implement local development activities. The methods to be employed will (i) permit

targeted communes and villages to establish or, where already in place, strengthen Village

Development Committees (VDCs) and Commune Development Committees (CDCs); (ii) reinforce

the capacities of de-centralized line ministry staff and local administrations to support the

planning and implementation of these activities; and (iii) foster trust and mitigate conflicts

between communities and Government and stimulate the effective communication between

diverse stakeholders involved in local development.

2. Component B: Priority Response Fund. The PRF’s objective is two-fold. It aims to (i) offer a

window of opportunity to expedite the delivery of most-needed resources; and (ii) encourage

participation and buy-in of the project by demonstrating a rapid tangible dividend to

collaboration. This component will disburse funds in small increments to finance sub-projects

that are ready for implementation. The PRF allows for rapid centralized project selection and

approval through a set of procedures that are lighter and more rapid than those under

component 3. As such, it will deliver immediate tangible benefits while giving the project time to

develop some of its more comprehensive but labor-intensive institutional structures.

3. Component C: Local Development Fund. The objective of the LDF is for targeted communes and

villages to have access to funding in a timely and transparent manner. The focus of the fund will

Page 16: Baseline Report for the Impact Evaluation of the Central ...

16 PDCAGV IE Baseline Survey Report

be on community public goods, with rural communes and villages identifying public socio-

economic investments through a local development planning process.

4. Component D: Project Management, Monitoring, and Evaluation. The objective of this

component is to effectively coordinate and monitor components A-C.

1.2 Research questions The IE was originally designed to answer three related research questions:

1. What is the impact of institutional capacity building under Component A? A serious challenge in

evaluating the impact of community-driven development (CDD) programs is separating the

impact of institutional development from the impact of actual service provision, and in

particular from the impact of budget transfers. At present, no rigorous IE exists that separates

the impact of local institutional development and identifies its importance for the functioning of

CDD programs and local development. The PDCAGV IE was, therefore, designed to allow for the

evaluation of the capacity-building component in isolation from the other project components.

2. How does the performance of centrally prescribed projects compare to the performance of

locally owned projects chosen through participatory processes? Typically, IEs of CDD projects

compare the outcomes in project areas with outcomes in suitably chosen control areas that are

not covered by the project. However, the most common alternative to CDD is not the absence of

any intervention but rather centrally managed development. Therefore the typical impact

evaluation of a CDD project fails to measure the true marginal effect of such a project.

The PDCAGV IE was designed to enable the measurement of the marginal effect of the

community-driven component of the project relative to the effect of the centrally managed

component. Similarly, the IE was to provide evidence on the complementarity between

centrally and locally managed interventions, specifically whether the implementation of the

centrally managed component encourages participation and generally improves the outcomes

of the locally managed intervention.

3. What are the overall effects of the project components on households, villages, and communes?

At the household level, the IE will measure the causal impact of the different project

components on outcomes related to income, consumption, nutrition, literacy, school

attendance, disease occurrence, social capital, and empowerment. At the village level, the IE

will measure the causal impact of the different project components on outcomes relating to

infrastructure availability and quality, availability of information, and responsiveness and

accountability of village leadership. At the commune level, the IE will measure the causal impact

of the different project components on outcomes related to conflict resolution and the inclusion

of women in commune leadership.

Page 17: Baseline Report for the Impact Evaluation of the Central ...

17 PDCAGV IE Baseline Survey Report

As the project will begin with the rehabilitation of water pumps under the PRF, the IE will likewise focus

initially on the impact of this component (question 3). As such the focus of this baseline report is on

household, village, and commune-level outcomes which may be important to assessing the impact

especially of this component. That said, the IE design remains in place to answer all three research

questions, contingent on developments in project implementation.

1.3 Impact evaluation design The PDCAGV IE uses a randomized roll-out design. Ten of CAR’s sixteen prefectures are participating in

the project,1 and all 102 communes in these prefectures were randomly assigned to one of four groups.

This was done through a lottery in which all communes had the same chance of being selected for each

group.

Initially, a three-phased evaluation design was proposed, with the four groups receiving different

components and combinations of components across time:

Group 1 (48 communes) was to receive only the capacity building component, starting in phase

one.

Group 2 (18 communes) was to serve as control group in phase 1, i.e. receive no intervention.

In phase two, Group 2 was to receive the capacity building component and, in phase three, the

LDF.

Group 3 (18 communes) was to receive the capacity building and PRF components in phase one

and, in phase two, the LDF.

Group 4 (18 communes) was to receive the capacity building and PRF components in phase one.

This is summarized in Table 1.3.1: Phasing of project components by evaluation group.

Table 1.3.1: Phasing of project components by evaluation group

Interventions

during Phase

Group (number of communes in group)

1 (48) 2 (18) 3 (18) 4 (18)

One (P1) Capacity building - Capacity building +

PRF

Capacity building +

PRF

Two (P2) - Capacity building LDF

-

Three (P3) - LDF - -

1 The prefectures are Basse Kotto, Kemo, Lobaye, Mambéré-Kadei, Mbomou, Nana-Mambéré, Ombella-M’Poko,

Ouaka, Ouham (Bossangoa sub-prefecture only), and Sangha-Mbaéré.

Page 18: Baseline Report for the Impact Evaluation of the Central ...

18 PDCAGV IE Baseline Survey Report

Communes within each of the ten project prefectures were randomly assigned to each of the four study

groups according to the following rules:

1. The total number of communes receiving capacity building as well as the PRF and/or LDF was

limited to 54 due to limited project resources;

2. Within every prefecture, there was to be a balance between the number of communes that

receive only capacity building and those that also receive PRF and/or LDF, i.e. in each prefecture

one group of communes should receive only capacity building, and three identically-sized groups

of communes should receive some combination of capacity building plus PRF and/or LDF;

3. The number of communes in each group should be proportional across prefectures.

The procedure for randomly allocating communes into each of the four study groups is described in

more detail in Appendix 1 of the concept note (available by request).2

Capacity building and the LDF components are commune level interventions. The PRF, however, is

implemented at the level of the village. Given the possibility that PRF resources might be insufficient to

cover basic infrastructure needs in all target villages (those within communes randomly allocated to

either Group 3 or Group 4) 3 and the nonexistent or low-quality data on village-level infrastructure, an

infrastructure needs assessment survey was carried out in August 2010.

The PRF component targets villages near main roads with a population greater than 200, and so villages

that did not meet these conditions were excluded from the project by default. From remaining villages

in each of the 36 PRF communes, a maximum of ten villages per commune was randomly selected and

surveyed. Data was collected on existing infrastructure in each of these villages, including variables such

as physical condition, age, previous renovations, rehabilitation costs and frequency of use. Data on

village-level infrastructure priorities was also collected. At the time of this survey, it was understood that

the project would favor villages with a school, thus all such villages had a greater probability of being

selected for the survey.

Data was collected on 243 of 309 selected villages in 34 communes.4 In the remaining 66 villages, there

was no public infrastructure to be rehabilitated, and as the PRF does not finance the construction of

new infrastructure, priority infrastructure for rehabilitation was identified in neighboring villages. The

survey was therefore carried out in these villages, so that rehabilitation of priority public infrastructure

could take place in these villages which was assumed to eventually benefit also the originally selected

villages due to geographic proximity. This was viewed as a fair and pragmatic way to provide the

possibility of benefitting from the PRF to those villages selected for the PRF component which were

subsequently found not to have any infrastructure to rehabilitate.

2 Please contact Marcus Holmlund ([email protected]) or see the project folder in WBDocs (P120087).

3 Given the overwhelming need for basic infrastructure at the village level, project resources initially allocated to

the PRF were indeed found to be inadequate. The project will therefore be restructured to increase the amount of funding available for the PRF. 4 Two communes, Nadziboro and Kounago, were inaccessible.

Page 19: Baseline Report for the Impact Evaluation of the Central ...

19 PDCAGV IE Baseline Survey Report

On the basis of data collected during the infrastructure needs survey, PDCAGV regional experts

identified one piece of priority infrastructure for rehabilitation in each surveyed village. In order to

determine relative need, a “PRF index” was then constructed on the basis of the following variables,

chosen to be in line with PRF objectives:5

1. Condition of infrastructure (rated on a scale of 1 to 6, with one corresponding to excellent and

six corresponding to very poor/unusable)

2. Frequency of use, on the condition that the infrastructure was still in usable condition

3. Age of infrastructure

4. Time since last reparations

5. Number of PRF-eligible villages in the commune

6. Presence of pygmies (rated on the basis of an index that combines three questions on pygmies

and runs from 0 to 8)6

Villages received a score for each of these variables, which was standardized by the total number of

potential scores for each variable. The final PRF index then simply sums the scores across variables for

each village. This produced a priority ranking of villages, which was discussed with regional officers and

survey teams and, with one exception, fit perceptions of which villages and infrastructures should be

prioritized (this is perhaps not surprising as target villages are quite homogenous in terms of

infrastructure availability/needs, poverty, and socio-demographic characteristics, so in essence every

village examined could be considered a priority). Given the relatively uniform distribution of

infrastructure needs, random allocation of priority to some villages would have been equally fair, but

was judged not to be politically feasible. Instead, the index was created to fulfill the need for PRF

allocation to be based on objective, observable characteristics in line with project goals.

This index was used to create ranking of villages by infrastructure rehabilitation need. Following this, a

cost threshold was defined to balance (i) the need for infrastructure rehabilitation under the PRF with

(ii) project resources available for this component.

In calculating the threshold/number of villages to benefit from the PRF, the maximum cost of any single

infrastructure rehabilitation was first set at CFA 20 million (about USD 40 thousand). If the village

infrastructure needs survey had defined a first priority infrastructure with a rehabilitation cost

exceeding this, this was automatically excluded and the second priority was instead considered, and so

on. If a village lacked an infrastructure rehabilitation cost estimate, the average from the survey as a

whole was used.

5 Several weighting schemes and index calculations were discussed with the project team, and the one described

here was selected by them as the most intuitive and transparent. 6 The project aims to benefit particularly vulnerable groups such as pygmies.

Page 20: Baseline Report for the Impact Evaluation of the Central ...

20 PDCAGV IE Baseline Survey Report

Of the 66 villages which did not have any public infrastructure to be rehabilitated, half were randomly

selected, and of the other 243 villages the cutoff value of the index was set at 3.22, meaning that all the

147 villages with an index value higher than this were selected to benefit from the PRF. A total of 180

villages were thus chosen for the PRF (list available by request).7

1.4 Data collection Baseline data was collected at the commune, village, and household levels using five questionnaires:

household, village, village chief, commune, and commune council (survey questionnaires available by

request).8

Of the 102 communes in the study area, a total of 72 were targeted, that is 18 communes selected

randomly from Group 1 and all communes in Groups 2 through 4. The commune council and commune

mayor surveys were to cover all of these. 357 villages were selected for the survey, and a total of ten

households were randomly sampled per village (that of the village chief and nine randomly selected

households). Table 1.4.1 summarizes the planned sample for each survey type.

Table 1.4.1: Target sample by survey type

Household

(including

village chief

household)

Village chief

household only

Village Commune

mayor

Commune

council

Target sample

size

3,570 357 357 72 72

Survey fieldwork was carried out in September and October of 2010. This was complicated by the poor

state of infrastructure and security in CAR, resulting in a final sample smaller than what was originally

envisioned. The following impeded survey implementation and data availability:

One survey team experienced a breakdown of their vehicle, which left them stranded for a

week. Nine villages were therefore dropped from the survey.

Security conditions precluded access to a five villages.

For non systematic reasons, a further fifteen villages could not be reached. This was primarily

due to extremely poor infrastructure (impassable roads or collapsed bridges). It is possible,

however, that certain questionnaires were lost between data collection in the field and data

entry in Bangui, and that village and commune-level questionnaires were not administered in all

cases.

7 Please contact Marcus Holmlund ([email protected]) or see the project folder in WBDocs (P120087).

8 As above.

Page 21: Baseline Report for the Impact Evaluation of the Central ...

21 PDCAGV IE Baseline Survey Report

The process of data entry was also significantly delayed, and the evaluation team will work to ensure

lessons learned from this experience result in a more efficient turnaround during follow-up data

collection. Nonetheless, the ICASEES staff noted in the Acknowledgements part of this report should be

recognized for their great efforts in delivering a final dataset which is, in many respects, almost

complete.

Table 1.4.2 compares the target sample by survey type to the actual sample observed in the datasets

Table 1.4.2: Target and final sample by survey type

Household (including

village chief household)

Village chief

household

only

Village Commune

mayor

Commune

council

Target sample size 3,570 357 357 72 72

Final sample size 3,201 324 324 64 64

Completion rate 90% 91% 91% 89% 89%

2. Baseline survey coverage This section provides a summary of the geographic coverage of the baseline survey and the final sample

sizes and geographic distribution for each of the five surveys. It should be noted that the data

presented in this report is only representative for PDCAGV project areas, and should not be interpreted

as representative of the Central African Republic as a whole.

2.1 Geographic coverage The data in this report cover the ten project prefectures. Figure 2.1.1 shows these prefectures, while

Table 2.1.1 summarizes the number of communes, villages, and households surveyed in each prefecture.

Figure 2.1.1: Geographic Coverage

Page 22: Baseline Report for the Impact Evaluation of the Central ...

22 PDCAGV IE Baseline Survey Report

Table 2.1.1: Geographic Coverage

Region Prefecture Commune

Surveys

Village

Surveys

Household

surveys

Region 1 Ombella-M'Poko 4 19 188

Region 1 Lobaye 7 34 337

Region 2 Mambere Kadei 7 36 364

Region 2 Nana-Mambere 12 58 571

Region 2 Sangha-Mbaere 5 21 206

Region 3 Ouham 3 20 186

Region 4 Kemo 4 20 191

Region 4 Ouaka 6 41 402

Region 6 Basse-Kotto 8 37 370

Region 6 Mbomou 8 38 386

TOTAL 64 324 3201

3. Household and individual-level results

3.1 Household size, age, and gender distribution The average household size is 5.25 (standard deviation: 3.1) and exactly 50 percent of surveyed

household members are male. The average age of household members is 20.7 years (standard

deviation: 17.7).

Table 3.1.1: Household demographics

Average household size 5.25

Fraction of male household members in the sample 50%

Average age of household member 20.7

Source: Household survey

Page 23: Baseline Report for the Impact Evaluation of the Central ...

23 PDCAGV IE Baseline Survey Report

Figure 3.1.1 below shows population over the course of the last decades. CAR’s population has doubled

since 1980 and now stands at an estimated 4.4 million people. Results from the follow up survey will

allow us to assess the rate of population growth in PDCAGV project areas versus that in the country as a

whole.

Figure 3.1.1: Population (in millions)

Source: Africa Development Indicators

Average per capita income of surveyed individuals was just over USD 19, with 73% of persons living

below the national poverty line. This is above the national average of 62% reported in the Africa

Development Indicators for 2008. Figure 3.1.2 shows the progression of the poverty rate according to

ADI and compares this to the rate estimated through the baseline survey (note that the figure for 2010

is for the baseline survey area only, thus the comparison to the earlier national poverty levels should be

made with this in mind).

Figure 3.1.2: Poverty Rates

1.5

1.8 2.2

2.9

3.6

4.4

1,25

1,75

2,25

2,75

3,25

3,75

4,25

4,75

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Po

pu

lati

on

(m

illio

n)

Page 24: Baseline Report for the Impact Evaluation of the Central ...

24 PDCAGV IE Baseline Survey Report

Source: Household survey and Africa Development Indicators

83,2

62,4 62,8 73,0

0

20

40

60

80

100

1992 2003 2008 Baseline Survey(2010)

Figure 3.1.2: Share of population living on less than $1.25 a day & sample data (national poverty line)

1992 2003 2008 Baseline Survey (2010)

Page 25: Baseline Report for the Impact Evaluation of the Central ...

25 PDCAGV IE Baseline Survey Report

3.2 Principal activities of household members (overall and by gender) Figure 3.2.1 illustrates the principal activities of household members (6 years of age and older) overall

and disaggregated by gender. Agriculture is by far the predominant primary activity: 52.8% of the

sample list agriculture as their primary activity. Of the men, 57.2% percent work in agriculture whereas

for women this ratio is 48.2%. 27.1% of the sample population are students (32.6% of women, 21.7% of

men). 8.1% of the sample is unemployed.

Figure 3.2.1: Principal household-member activities

Source: Household survey

52,8

27,1

8,1

3,3

2,6

1,9

1,8

0,6

0,6

0,5

0,4

0,2

0,0

57,2

21,7

9,0

5,6

2,7

1,0

2,1

0,0

0,1

0,2

0,2

0,1

0,0

48,2

32,6

7,1

1,0

2,6

2,7

1,6

1,2

1,0

0,9

0,7

0,4

0,0

0 10 20 30 40 50 60

Agriculture

Student

Unemployed

Household tasks

doesn't apply (too old/young)

Mining

Services/small business

Hunting

Private employee

Other

Public employee

Handcraft

Soldier

Figure 3.2.1: Principal activity of household members overall & by gender (% of household members reporting principal activity), n = 13,170

Total Male Female

Page 26: Baseline Report for the Impact Evaluation of the Central ...

26 PDCAGV IE Baseline Survey Report

For CAR as a whole, figure 3.2.2 shows that, while still 65% of the population is dependent primarily on

agricultural income, this number is down from almost 100% in 1960. Currently, across sub-Saharan

Africa 55% of the population is primarily dependent on income from agriculture.

Figure 3.2.2: Agricultural dependency

Source: Africa Development Indicators

2.8 (65%)

4.3

1

1,5

2

2,5

3

3,5

4

4,5

5

1960 1968 1976 1984 1992 2000 2008

Mill

ion

s

Figure 3.2.2: Agricultural dependency

Agricultural population Population

Page 27: Baseline Report for the Impact Evaluation of the Central ...

27 PDCAGV IE Baseline Survey Report

3.3 Illnesses preventing work/going to school In the month preceding the survey, 26% of household members were unable to go to school or work

due to illness. Table 3.3.1 shows the primary reasons for absence from work/school.

Table 3.3.1: Reasons for absence from work

Diarrhea 21.1

Malaria 13.4

Fever 12.5

Flu 12.2

Parasites 7.1

Injury 4.7

Hypertension 3.1

Skin condition 1.9

Anemia/malnutrition 1.1

Diabetes 0.7

Other 22.3

Source: Household survey

3.4 Education status of household members In order to assess the education level of household members, we use proxies such as literacy in French

and school attendance. In addition, table 3.4.1 also includes the percentage of household heads that

have had at least elementary education. The data shows that 31% of the sample is able to speak French,

CAR’s official language. The proportion is higher among men (47%) than for women (16%). Literacy

levels are lower in the sample when compared to country-wide data from the World Bank’s Africa

Development Indicators (ADI) database, which indicate that 55% of the population is literate (69% of

men and 42% of women; figures for 2010).

62% household members between the ages of 6-14 have attended school at some point (67% of males,

56% of females), and 57% were attending school at the time of the survey (63% of males, to 51% of

females).

Page 28: Baseline Report for the Impact Evaluation of the Central ...

28 PDCAGV IE Baseline Survey Report

Table 3.4.1: Education indicators

Sample

Country

(ADI)

a. Adult Literacy 31% 55%

Male 47% 69%

Female 16% 42%

b. Whether or not household member between 6-14 has ever attended school 62%

Male 67%

Female 56%

c. Whether or not household member between 6-14 currently attends school 57%

Male 63%

Female 51%

d. Household head has at least elementary education (CEM2 and above) 32%

Source: Household survey, Africa Development Indicators

Table 3.4.1 illustrates that, at the national level, CAR experienced a drastic improvement in primary

school enrollment rates. While numbers hovered around 50% for the last decades, school enrollment

leaped by 21 percentage points in between 2005 and 2010, from 48% to 69%. Results from the follow up

survey will allow us to assess changes in school enrollment in PDCAGV project areas and to compare

these to results for the country as a whole.

Figure 3.4.1: Net primary school enrollment

50 55 53

48

69

0

10

20

30

40

50

60

70

80

1971 1981 1991 2005 2010

Figure 3.4.1: CAR net primary school enrollment (%)

Source: Africa Development Indicators

Page 29: Baseline Report for the Impact Evaluation of the Central ...

29 PDCAGV IE Baseline Survey Report

3.5 Health/nutrition indicators The baseline household survey revealed high rates of stunting and malnutrition among children under 5

years of age: 61% are stunted and 64% suffer from moderate or severe malnutrition. Stunting9 and

malnutrition are measured in terms of height-for-age and weight-for-age, and are generally considered

to be relatively reliable indicators of household welfare as measures of household income and

consumption often suffer from recall bias.10 Furthermore, in rural areas where many households are

engaged in subsistence farming income and expenditure may not be adequate or appropriate welfare

measures.

The very high rate of stunting and malnutrition in children under 5 is a cause for serious concern, as in

addition to direct health impacts these may retard cognitive and physical development. Given findings

on daily nutrition, however, they are not surprising. The average number of meals for surveyed

households was only 1.6 per day (standard deviation: 0.77), and merely 78% percent of these were

reported as being filling/satisfactory (the equivalent of 1.2 filling/satisfactory meals per household per

day).

Interestingly, the extent of moderate and severe malnutrition found in the PDCAGV baseline survey

sample is significantly higher than the national figure reported by ADI (22% in 2000). This may warrant

further investigation.

Table 3.5.1: Selected health indicators

sample ADI (2000)

Moderate & severe stunting of children under 5 61%

Moderate & severe malnutrition of children under 5 64% 22%

Avg. number of meals on day preceding survey 1.6

Fraction of these which were filling/satisfactory 78%

Source: Household survey

Though comparisons between the PDCAGV baseline and national-level figures from ADI should be

interpreted with caution, it is instructive to note that the under 5 mortality rate, as shown in figure

3.5.1, is currently higher in CAR than it is across the region as a whole. Poor nutrition is likely a

contributing factor to this high mortality rate.

9 Stunting is defined as having a height (or length)-for-age more than two standard deviations below the median of

the WHO growth reference. It is calculated by taking body measurements of height or length. Other data needed are age and gender. 10

Recall bias occurs when the answer to a question is affected by the respondent’s memory.

Page 30: Baseline Report for the Impact Evaluation of the Central ...

30 PDCAGV IE Baseline Survey Report

Figure 3.5.1: Under-5 mortality rate

Source: Africa Development Indicators

3.6 Type of dwelling 93% of respondents own their dwelling, and Figure 3.6.1 illustrates the types of dwelling found in the

study areas. The majority (58.4%) of households live in an individual house whereas 22.7% live in a hut.

8.6% possess multiple houses, 5.9% own multiple huts, and 3.6 %% have both multiple houses and huts.

Only 0.6 % of the surveyed households live in an apartment.

Figure 3.6.1: Type of dwelling

Source: Household survey

Figures 3.6.2 and 3.6.3 below show the prevalence of traditional materials in the construction of

dwellings. 59.3% of roofs are made from thatch, with bamboo representing another 22.3%. The

prevalence of thatch roofs is fairly consistent with national results from the “Enquête Centrafricaine

pour le Suivi et Evaluation du Bien-être“ (ECASEB) survey from 2008. However, Figure 3.6.2 shows that

58,4

22,7

8,6 5,9 3,6 0,6 0,2

IndividualHouse

Hut Multiplehouses

Multiple huts Huts andhouses

Apartment Other

Figure 3.6.1: Type of dwelling (%)

226

130

317

175 171

50

100

150

200

250

300

350

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2009

Figure x: Under- 5 mortality rate (per 1,000 births) in CAR and SSA

SSA CAR

Page 31: Baseline Report for the Impact Evaluation of the Central ...

31 PDCAGV IE Baseline Survey Report

nation-wide, more roofs are made of tin rather than of bamboo as in the PDCAGV baseline sample.

(16.3% tin roofs in the sample, compared to 22.3% nation-wide). As shown in figure 3.6.3, 78% of walls

are made from fired bricks, while non-fired bricks make up another 11%.

Figure 3.6.2: Roof material

Source: Household survey and ECASEB

Figure 3.6.3: Wall material

Source: Household survey

59,8

12,4

26,8

0,1

0,6

0,2

59,3

22,3

16,3

1,4

0,4

0,2

thatch

bamboo

tin

bricks

other

concrete

Figure 3.6.2: Roof material (% of roofs made from)

ECASEB Survey

78%

11%

5% 5% 1%

Figure 3.6.3: Main wall material

fired bricks non-fired bricks banco mud-brick other concrete bricks

Page 32: Baseline Report for the Impact Evaluation of the Central ...

32 PDCAGV IE Baseline Survey Report

3.7 Time to access key infrastructure Table 3.71 displays the mean and median time for villagers to access certain points that are vital for

their daily routines. While a water point is on average 30 minutes away from the house and a primary

school can be accessed in 45 minutes, other facilities are further out: It takes more than 90 minutes to

reach a health facility, a spot with access to electricity, or an area with cellular coverage (93, 107, 98

minutes respectively on average). Reaching the market is even harder: it typically takes 159 minutes to

get there.

Table 3.7.1: Time to access locations of interest (average by foot in minutes)

Access by foot in minutes Mean Median

Water point 30 10

Electricity 107 45

Mobile phone network 98 10

Primary school 45 20

Health facility 93 60

Market 159 120

Source: Household survey

Page 33: Baseline Report for the Impact Evaluation of the Central ...

33 PDCAGV IE Baseline Survey Report

3.8 Ethnicity/religion Baseline survey results show a variety of ethnicities and religious affiliations in the baseline study areas.

The largest ethnic group is the G’baya (37.5% of households). The second largest ethnic group is Banda,

which represents 22.4% of the surveyed population. With more than 40% of households belonging to

other ethnic groups, the survey highlights the ethnic diversity of CAR.

Figures from the PDCAGV baseline are comparable to national statistics, derived from the CIA’s world

fact book. However, the sample contains less people stemming from the Mandja, Sara, and Mbaoum

than the national average, suggesting that these ethnicities do not have strong populations in the

surveyed areas. These discrepancies are illustrated in figure 3.8.1

Figure 3.8.1: Ethnicity

Sources: Household survey and CIA World Fact Book

37,5

28,2

22,4

5,4 3,7

2,1 0,4 0,3

33

2

27

13

4 4

10 7

G'baya Other Banda Mandja G'baka Yakoma Sara Mbaoum

Figure 3.8.1: Ethnicity (% of population)

Survey Country (CIA World Fact Book)

Page 34: Baseline Report for the Impact Evaluation of the Central ...

34 PDCAGV IE Baseline Survey Report

Figure 3.8.2: Mother tongue

Source: Household survey

While most people are Christian (54.1% Protestant, 32.1% Catholic), there is a 6.4% Muslim minority.

Figure 3.8.3: Religious affiliation

Source: Household survey

46,2%

37,5%

11,4%

4,9%

Sango Other Gbaya Banda

Figure 3.8.2: Mother tongue of respondents

54,1%

32,1%

6,4% 5,5% 1,3% 0,6%

Protestant Catholic Muslim Other Animist No religion

Figure 3.7: Religious affiliation (%)

Page 35: Baseline Report for the Impact Evaluation of the Central ...

35 PDCAGV IE Baseline Survey Report

3.9 Perception of village problems

Almost a third of the people (30.5%) identified insufficient access to drinking water as the main problem

of their village. Absence of hospitals (23.4%) and absence of schools (17.0%) were the second and third

most common problems cited. 9.5% of respondents cited the inferior conditions or lack of roads.

Figure 3.9.1: Village problems

Source: Household survey

30,5

23,4

17,0

9,5

4,1

3,8

2,8

2,7

2,3

1,9

0,8

0,8

0,5

Access to drinking water

Lack of hospital

Lack of school

Roads

Other

Market access

Lack of doctors

Jobs

Agricultural tools

Lack of teachers

Electricity

Security/Violence

Irrigation

Figure 3.8: Most important village problems (% of household heads reporting problem)

Page 36: Baseline Report for the Impact Evaluation of the Central ...

36 PDCAGV IE Baseline Survey Report

3.10 Prevalence of theft, violence, and conflict Table 3.10.1 shows prevalence of theft and violence in the village. 25% of respondents report theft of

their animals over the last 6 months. Another 18% say that personal goods have been stolen in that time

period. Violence between village inhabitants was observed by 16% of the surveyed population while

violent acts by outsiders were noticed by 11% of the sample. 6% of respondents state that either they or

another household member had been a victim of a violent act in the past six months.

Table 3.10.1: Prevalence of violence and theft

Theft and violence over the last 6 months – respondents reporting: %

Theft of respondent's animals 25

Theft of respondent's goods 18

Violence between village inhabitants 16

Violence by people from outside the village 11

Respondent was victim of a violent act 6

Household member was victim to violence 6

Whether a family member has been injured/killed due to conflict in past 3 years 7

Whether anyone from the village has been injured/killed due to conflict in past 3 years 11

Whether a person was reported as missing/displaced 18

Whether the village has a self-defense committee (militia) 67

Source: Household survey

Page 37: Baseline Report for the Impact Evaluation of the Central ...

37 PDCAGV IE Baseline Survey Report

3.11 Personal security The vast majority (88%) of respondents say that they live in a relatively quiet village, with 87% and 85%

reporting that they feel safe walking in the village during the day and night, respectively. 80% feel safe

walking outside the village and 59% report not being afraid of violence and crime when they are home

alone. 35% have never heard firearms in their village, and only 9% agree that their village is marked by

violence, with 6% stating that the level of violence has risen the last 2 years and that it is very likely they

will be become a victim of violence in the upcoming year. 34% percent respondents avoid certain roads

and areas where they don’t feel safe enough.

Table 3.11.1: Personal security perceptions

Respondents agreeing with the following statements: %

I live in a quiet village 88

I feel safe walking in the village during the day 87

I feel safe walking in the village during the night 85

I feel safe walking outside the village 80

I'm not afraid of violence and crime when I am alone at home 59

I never heard fire-arms in my village 35

I avoid certain roads and don't enter territory where I don't feel secure 34

My village is marked by prevalence of violence 9

In my village occurrence of violence has risen during the past 2 years 6

It is very likely that I will become a victim of violence during the coming 12 months 6

Source: Household Survey

3.12 Participation in village meetings/volunteerism More than half (59%) of the households had a member participating in a village meeting in the last 12

months and 31% had a household member speak in such a meeting. Reported willingness to volunteer

for the community, at 85% percent, is high among villagers.

Table 3.12.1: Participation in village affairs (%)

Households reporting

That a household member has participated in a village meeting in the past 12 months

(conditional on such a meeting having taken place) 59%

That a household member has spoken at a village meeting in the past 12 months

(conditional on such a meeting having taken place) 31%

Willingness to volunteer for the benefit of the community 85%

Source: Household Survey

Page 38: Baseline Report for the Impact Evaluation of the Central ...

38 PDCAGV IE Baseline Survey Report

3.13 Trust Trust within households, villages, and communities is an important component of social capital and an

indication of social welfare. Project activities, particularly aspects of the capacity building and LDF

components, are expected to increase trust at all levels, in addition to themselves being influenced by

pre-existing levels of trust. Household survey respondents reported that, on average, they feel

comfortable speaking with 2.14 people when facing personal problems. The average number of people

that could lend the household 10,000 CFA on short-notice is 1.14.

It is striking that, when asked who they would go to for help in case their mobile phone gets stolen,

67.6% of respondents site the village chief, illustrating the importance of traditional systems of

governance. The next most popular answer, selected by 9.4% of respondents, is to not ask for help.

Table 3.13.1: Who would you ask for help in case your mobile phone gets stolen? (% of respondents

citing option)

Village chief 67.6%

Person doesn't ask for help 9.4%

Gendarmerie 5.8%

Other 5.3%

Person goes after the criminal him-/herself 4.7%

Mayor of the commune 3.9%

Family/Clan 1.1%

Police 1.0%

Commune council 0.5%

Neighbors 0.4%

Military 0.3%

Source: Household Survey

3.14 Attitudes Table 3.14.1 reports the answers to attitude questions. 99% agree that leaders should be selected by

means of free regulated and honest elections, and some 93% believe that youths can serve as a

respectable local leader. 87% think that women can be good leaders and must be encouraged to assume

public office.

Page 39: Baseline Report for the Impact Evaluation of the Central ...

39 PDCAGV IE Baseline Survey Report

90% of respondents agree with the statement that people behave like children, and that the

government needs to take care of them. More than half (51%) agree that elected leaders should

represent everyone, not solely their own constituency. Less than half the survey respondents (42%)

denounce violence as a just mean to pursue political goals and 29% believe that it is acceptable for a

husband to beat his wife when she behaves badly. 17% of the sampled population agrees that it is

normal to bribe an official in order to speed up work.

Table 3.14.1: Personal attitudes

Do you agree with the following statement? %

We need to choose our leaders by means of free, regulated and honest elections 99

A responsible youth can be a respectable local leader 93

The people are like children, the government needs to take care of them 90

Women can be good leaders, they need to be encouraged to assume office 87

Since leaders represent everyone, they should not favor their own family or group 51

Use of violence is never justified by following political goals 42

A husband can beat his wife if she behaves badly 29

In our country, it is normal to bribe an official to help speed up work 17

Source: Household Survey

Page 40: Baseline Report for the Impact Evaluation of the Central ...

40 PDCAGV IE Baseline Survey Report

3.15 Nganga contacts/supernatural powers Supernatural powers and Nganga (spiritual healers) play an important role in CAR. 7% of the survey

respondents reported that they possessed supernatural powers, and 17% of households have consulted

Nganga in the past. Of those, 64.1% did so in order to prevent/cure an illness, 15.3% in order to cure

poisoning, and 10.9% to achieve personal success.

Table 3.15.1: Nganga contacts/supernatural powers (in %)

Proportion of respondents believing they possess supernatural powers 7%

Proportion of households that consulted Nganga 17%

Reasons for consulting Nganga %

Illness 64.1

Poisoning 15.3

Personal success 10.9

Other 4.8

Success of harvest 3.1

Fertility 0.9

Success of children 0.6

Argument with someone from the village 0.2

Argument with someone from another village 0.2

Source: Household Survey

Page 41: Baseline Report for the Impact Evaluation of the Central ...

41 PDCAGV IE Baseline Survey Report

4. Village-level results

4.1 Village size Overall, 324 villages were surveyed. Within these villages, an average 9.9 households were surveyed

from each village (total: 3,201). The households contained an average of 5.25 people (standard

deviation: 3.11). Information on close to 17,000 people was thus captured by the baseline survey.

4.2 Village access Figure 4.1.1 shows the type of main access road to the village. 53.1% of surveyed villages were

accessible via a national road. Another 36.4% could be accessed by a less frequented road. Mobile

phone networks were accessible only in 27% of the surveyed villages, which speaks to the relative

isolation of an important proportion of project areas. Table 4.1 indicates the state that the main road to

the village is in.

Table 4.1.1: Condition of main village access road

Proportion of villages with access roads that are Accesible by vehicle 36.4

Difficult to pass by car (bad state) 28.4

Recently repaired (good state) 19.8

Impassible by car (4x4) 10.2

No road available 4.0

N/A 1.2

Source: Village survey

Figure 4.1.1: Village access

Source: Village survey

53,1

36,4

5,6 4,0 0,9

N ational road Earth road Other Footpath Canoe

Figure 4.1.1: Main way to access village (% of villages)

Page 42: Baseline Report for the Impact Evaluation of the Central ...

42 PDCAGV IE Baseline Survey Report

4.3 Pygmy populations around villages Figure 4.2 illustrates the prevalence of pygmy populations in and around the sampled villages, and show

clearly that pygmies represent a small minority of the population in the study area. The vast majority of

villages (91.7%) have no pygmy population; 6.8% of surveyed villages have a population of 0-50 pygmies;

and only 1.5% of villages have a pygmy population greater than 50.

Figure 4.2.1: Pygmy population

Source: Village survey

91,7

4,0 2,8 1,5

88,6

4,9 3,4 2,2

89,2

6,5 2,2 1,2

None 0-20 21-50 51 and more

Figure 4.2.1: Pygmy population (% of villages with pygmies)

Living in village Living close to village Employed in the field

Page 43: Baseline Report for the Impact Evaluation of the Central ...

43 PDCAGV IE Baseline Survey Report

4.4 Availability and condition of key infrastructure in villages Figure 4.4.1 shows key infrastructure availability in sampled villages. The most common village

infrastructure includes water pumps (38.9%), primary schools (33.6%), and mobile phone networks

(27%). Almost none of the surveyed villages have a secondary school (0.3%). Health facilities (12.3%) and

larger health centers (2.8%) are also rare. Generally speaking, the situation vis-à-vis basic health,

education, and sanitation infrastructure at the village level is extremely poor.

Figure 4.4.1: Key infrastructure

Source: Village survey

Figure 4.4.2 describes the differences in perceived village infrastructure needs as expressed by (i) village

chiefs and (ii) survey teams. Both chiefs and survey teams rank primary schools as the top priority for

rehabilitation (i.e. it is included in their top three choices in 25.3% and 22.2% cases, respectively, more

than any other infrastructure type), and their construction need is similarly assessed to be relatively high

(21.9% of village chiefs and 18.5% of survey teams site this as one of the top three construction

priorities). Rankings for water pumps across chiefs and survey teams are also similar for both

construction and rehabilitation, although the need for the former is seen as slightly more pressing by

survey teams. Relatively low importance is accorded to depository warehouses, markets, fortified

water sources, sheltered wells, and kindergartens.

In the case of roads, both village chiefs and survey teams see a greater need for constructing new roads

as opposed to rehabilitating existing ones, which is consistent with the data on village accessibility

highlighted above. For health facilities and health centers, both chiefs and the survey team accord

higher priority to rehabilitating existing structures, suggesting that, while these may be physically

present in study areas, their current status is sub-optimal.

38,9

33,6

27

12,3 11,7

6,8 6,5 4,9 3,4 2,8 0,9 0,6 0,3 0,3

Figure 4.4.1: Availability of key infrastructure in the villages (% of villages with...)

Page 44: Baseline Report for the Impact Evaluation of the Central ...

44 PDCAGV IE Baseline Survey Report

Also, it is interesting to note that survey teams reported no infrastructure priorities for construction

and/or rehabilitation relatively more than did village chiefs (21.9% versus 17.0% for construction; 9.6%

versus 5.6% for rehabilitation). These results, though not surprising, are encouraging in that they

demonstrate that village chiefs are, in general, more aware of the infrastructure needs of their villages

than are outside observers. An alternate interpretation could be that village chiefs perceived the survey

as a potential means to get resources/projects for their village, and so over-reported construction and

rehabilitation needs. Unfortunately, without the benefit of expert external assessments we cannot

disentangle these two effects.

Page 45: Baseline Report for the Impact Evaluation of the Central ...

45 PDCAGV IE Baseline Survey Report

Figure 4.4.2: Infrastructure rehabilitation and construction needs

Source: Village survey

25,3

19,1

17,0

13,3

5,6

4,3

3,4

2,8

2,2

2,2

1,2

1,2

22,2

17,9

21,9

15,7

4,9

2,2

1,9

4,0

3,4

1,2

1,2

0,9

21,9

23,8

5,6

4,0

0,9

20,4

0,6

1,9

0,6

13,0

0,3

3,1

18,5

28,4

9,6

4,6

0,3

18,5

0,6

1,5

0,3

10,5

0,6

3,1

Primary school

Water pump

Missing or None

Road

Water source

Health facility

Depository warehouse

Market

Fortified source

Health center

Sheltered water well

Kindergarten

Figure 4.4.2: Infrastructure rehabilitation/construction needs according to village chiefs and teams (% of times that item appears in list of top three

priorities)

Rehabilitation need (Chief) Rehabilitation need (Team)

Construction need (Chief) Construction need (Team)

Page 46: Baseline Report for the Impact Evaluation of the Central ...

46 PDCAGV IE Baseline Survey Report

In the majority of villages (82.2%), the inhabitants use infrastructure of neighboring villages. Primary

schools are the form of infrastructure that is used the most in other villages. And this is consistent with

village chiefs’ and teams’ assessment that primary schools are a priority regarding

construction/rehabilitation of infrastructure (See Figure 4.4.2 above). Health facilities (16.4%) and larger

health centers (19.8%) are also among infrastructure accessed in neighboring villages. Water pumps are

used in 7.4% of cases.

Table 4.4.1: Primary types of infrastructure used in neighboring villages

% of villages using infrastructure in neighboring villages

Primary school 22.8

No use of other village's infrastructure 21.3

Health center 19.8

Health facility 16.4

Water pump 7.4

Market 6.2

Secondary school 2.8

Source (water) 1.2

Fortified source (water) 0.6

Kindergarten 0.6

Road 0.6

Access to electronic grid 0.3

Source: Village survey

Page 47: Baseline Report for the Impact Evaluation of the Central ...

47 PDCAGV IE Baseline Survey Report

5. Commune-level results Data was collected from 64 communes in the 10 study prefectures. The primary purpose of this data is

to highlight commune level indicators/phenomena which may play an important role in determining

village or commune level outcomes, either because they act as enabling factors or because spillovers

from commune-level activities may reach individual household and villages. Also, the collection of data

at the commune level will allow us to measure any commune-level impacts of the project.

5.1 Commune leadership The household survey results show that the population is split evenly between men and women.

Although 87% of household survey respondents think that women can be good local leaders and should

be encouraged to assume office (see section 3.14), 93% of commune council members are male.

5.2 Commune transparency Table 5.2.1 illustrates the openness of commune council meetings to the public. It shows that 90.3% of

the last council meetings were publically announced and that, in 83.9% percent of cases, the agenda was

made public before the council meeting. Minutes or decisions of the respective meeting were made

public for around three quarters of council meetings (75.8%), and in 56.5% of cases this information was

posted in public places. In addition, in 66.1% of communes surveyed the last budget was made public

and 60.7% revealed their administrative accounts. During the same period, however, the administrative

accounts were only requested an average of 2.2 times (standard deviation: 5.9), demonstrating rather

low demand for this form of openness.

Table 5.2.1: Commune councils and the public

% of communes where

the last commune council meeting was publically announced 90.3%

the agenda was made public before the last commune council meeting 83.9%

the minutes/decisions of the last commune council meeting were made public 75.8%

the minutes/decisions were posted/published in principal public places 56.5%

the last commune budget was made public 66.1%

administrative accounts were made public within the last 12 months 60.7%

Source: Commune survey

Page 48: Baseline Report for the Impact Evaluation of the Central ...

48 PDCAGV IE Baseline Survey Report

5.3 Commune Budget Figure 5.3.1 illustrates average commune budget evolution from 2008 to 2010. Average budget

decreased from 11.23 million FCFA in 2008 to 4.12 million FCFA in 2009. In 2010, budgets rose to 16.15

million FCFA. It is not immediately clear why budgets fell so dramatically (by 63%) between 2008 and

2009, and then increased by nearly 300% between 2009 and 2010.11

Figure 5.3.1: Commune budget

Source: Commune Survey

5.4 Commune tax revenues Figure 5.4.1 shows the proportion of communes (i) collecting various types of taxes and (ii) sharing tax

revenue with their prefecture. Close to ninety percent of communes collect taxes for civil registry

(90.6%) and market activity (89.1%). These tax revenues are shared with the prefecture in 41.8% and

42.6% percent of cases, respectively. The highest sharing ratios are for business and firearms taxes, with

57.1% and 53.6%, respectively, of communes collecting such taxes sharing them with their prefecture.

The figure shows the considerable variation across communes in terms of tax revenue collection and

sharing.

11

These budget figures should be interpreted with extreme caution, as they were not always well-recorded by enumerators and respondents sometimes gave what appeared to be inconsistent answers

16,15

4,12

11,23

201020092008

Figure 5.3.1: Commune budgets in 2010 (in million FCFA)

Page 49: Baseline Report for the Impact Evaluation of the Central ...

49 PDCAGV IE Baseline Survey Report

Figure 5.4.1: Tax collection and sharing with prefectures

Source: Commune survey

90,6 89,1

71,9

56,3 54,7

46,9 45,3

39,1

23,4 20,3 18,8 18,8

12,5

41,8 42,6

51,2

32,4

39,4 34,5

53,6

45,8 43,8

30,8

57,1

21,4

33,3

0

10

20

30

40

50

60

70

80

90

100

Figure 5.2: Types of taxes collected and shared with prefectures (% of communes)

% of commune sample (N=64) Revenue is shared with Prefecture in %

Page 50: Baseline Report for the Impact Evaluation of the Central ...

50 PDCAGV IE Baseline Survey Report

5.5 Principal agricultural products of communes Table 5.5.2 shows the principal agricultural products of the surveyed communes. The overwhelming

majority (67.2%) cultivate manioc, and around one fifth of the communes indicate that they raise coffee

beans (18.8%). Cotton, maize, and cultivation of other agricultural products add up to about 10% of total

agricultural outputs. This shows the low diversification of agricultural products in surveyed communes,

and raises concern over the general extent of food and livelihoods security in these areas.

Table 5.5.2: Principal agricultural product of communes

% of communes cultivating…

Manioc 67.2%

Coffee 18.8%

Cotton 4.7%

Maize 3.1%

Other 3.1%

Missing 3.1%

Source: Commune survey

Page 51: Baseline Report for the Impact Evaluation of the Central ...

51 PDCAGV IE Baseline Survey Report

5.6 Community organizations Figure 5.6.1 displays the prevalence of community organizations in the 64 surveyed communes. On

average, each commune features almost 86 community organizations. The types of organizations that

are most prevalent are religious organizations, parental organizations, and local development

associations with, on average, more than 10 of these organizations per commune. The high prevalence

of agricultural associations (9.25 on average) underscores the importance of agriculture in the economy

of CAR. Unions/syndicate structures outside of agriculture are rarer. Syndicate structures are almost

non-existent (0.25 per commune on average), and the same is true for merchant’s associations (1.24 per

commune on average).

Figure 5.6.1: Community organizations

Source: Household survey

10,60

10,30

10,08

9,25

8,03

6,61

5,22

5,17

4,84

4,68

4,48

3,75

2,93

2,22

1,24

0,64

0,25

Religious associations

Parental organizations

Local development associations

Agricultural co-operations

Youth associations

Sports clubs

Local political party committees

Women's groups

Villagers' financial associations

Other producers' associations

Cultural associations

Associations of water users

Traditional associations

Animal breeder's assocations

Merchants' associations

Other consumers' associations

Syndicate structures

Figure 5.6.1: Prevalance of groups & associations (average number of groups per commune)

Page 52: Baseline Report for the Impact Evaluation of the Central ...

52 PDCAGV IE Baseline Survey Report

5.7 National/international NGO presence in communes NGOs or other development partners are active in the majority of communes. 46% of communes have

either national NGO operations or some other national development program/project, while 53% of

communes have either international NGO operations or some other international development

program/project.

Though the presence of local/international NGOs and other development operations in more than half

of communes suggests a relatively high level of dependency on outside assistance, 44% of communes

surveyed indicated they have a local development plan suggesting that in many cases communes play,

or aspire to play, an active role in their own development.

Table 5.7.1: Presence of national NGOs, international development programs, local

development plans

Proportion of communes with:

Presence of national NGOs or development programs/projects 46%

Presence of international NGOs or development programs/projects 53%

A local development plan 44%

Source: Commune survey

The presence of national NGOs or development programs/projects correlates with lower involvement of

international NGOs/projects. Communes that currently have such a national presence have less

international involvement (42% of such communes have international projects, while 58% do not).

Conversely, communes without international NGOs or development programs/projects have more

national programs (59% with; 41% without).

Page 53: Baseline Report for the Impact Evaluation of the Central ...

53 PDCAGV IE Baseline Survey Report

6. Village Chief Survey The Village chief survey was administered to 317 chiefs from 10 prefectures. It was designed to capture

the roles and identities that chiefs ascribe to themselves, and to identify how they assumed office and

manage village affairs.

6.1 Basic summary statistics The following two tables show summary statistics for the village chief survey. As shown in table 6.1.1,

the vast majority of chiefs are male (94%). 45% of the chiefs report that their fathers were also chiefs, so

although a large fraction of chiefs have “inherited” their posts, this is by no means the rule. 14% also

serve as a traditional authority/chief and 16.4% implemented micro projects within the last 12 months.

Table 6.1.2 describes the average age of respondents as well as the length of their tenure. On average,

chiefs are 50.6 years of age and hold their position for 9.5 years. There is, however, considerable

variation around these averages.

Table 6.1.1: Village Chief Summary statistics

Proportion of respondents %

That are male 94.0 45.0 14.0 16.4

Whose father was village chief

That are also a traditional authority/chief

That have implemented micro projects in last 12 months

Source: Village Chief Survey

Table 6.2.1: Average age and years in office

Average Standard deviation Maximum Minimum

Age of respondent 50.6 12.5 66 29

Years in office 9.5 10.9 52 0

Source: Village Chief Survey

Page 54: Baseline Report for the Impact Evaluation of the Central ...

54 PDCAGV IE Baseline Survey Report

6.2 Ascent to power Figure 6.2.1 shows how the village chiefs assumed their posts. In more than 60% of cases, the chief was

elected, suggesting political competition exists at the local level (that said, this figure is self-reported, so

should be interpreted with caution). 19.6% claimed power due to traditional roles. Another 11.4% and

4.1% were replacements after death and resignation, respectively.

Figure 6.2.1: Assumption of power

Source: Village chief survey

60,9

19,6

11,4

4,1 1,3 0,6

Elected TraditionalRole

Replacementafter Death

Replacementafter

Resignation

SpecialDelegation

Other

Figure 6.1: How did village chiefs assume power? (% of village chiefs)

Page 55: Baseline Report for the Impact Evaluation of the Central ...

55 PDCAGV IE Baseline Survey Report

6.3 Chief roles The following graphs and tables illustrate the chiefs’ positions within their respective villages and

communities. Just over a third of chiefs (34.4%) are part of a local development organization. Some

32.5% belong to producer organizations, showing chiefs’ interests in economic matters, which is

somewhat out of synch with the relatively low importance accorded to commerce in the household

survey. 30% belong to women and youth groups, even though 94% percent of chiefs are men aged an

average of 50.6 years. As observed in the commune questionnaire, the village chief survey also suggests

relatively low importance of unions, with only 2.5% of chiefs belonging to a union.

Figure 6.3.1: Chiefs’ participation in community groups

Source: village chief survey

The vast majority (92%) of chiefs do not hold another administrative position in addition to their, and

only 7% are engaged at the commune level. Almost none of the chiefs surveyed belonged to a political

party, with 95.6% responding that they have no political affiliation. It is possible, however, that this

result is due in part to respondents’ reluctance to reveal political party affiliation to enumerators.

Table 6.3.1: Other administrative positions of village chiefs

Administrative position at… %

Does not hold position 92%

Commune Level 7%

Prefecture Level 1%

National Level 1%

Military 0%

Source: Village chief survey

34,4

32,5

30,0

19,6

18,0

14,5

11,0

7,3

2,5

Local Development Association

Producer Organization

Women and Youth Group

Local Political Comittee

Other

Sports Club

Cultural Group

Business Association

Union

Figure 6.2.1: Chiefs' participation in social groups (% of chiefs participating)

Page 56: Baseline Report for the Impact Evaluation of the Central ...

56 PDCAGV IE Baseline Survey Report

Table 6.3.2: Party membership of village chiefs (%)

Proportion of chiefs reporting affiliation to political party

Not Member 95.6%

KNK 3.8%

KNK (BDR) 0.3%

Source: Village chief survey

In general, village chiefs do not appear to be highly politically connected. As shown in Table 6.3.3, 76%

of respondents report not having relatives holding public office. A combined 20% stated that they have

relatives at national, commune, or prefecture levels that hold an administrative position and a small

group of chiefs reported to have relatives in the military (3%).

At the commune level, only 10% of respondents are members of the commune council. 2% and 1%,

respectively, are vice-president or president of their commune council (see figure 6.3.2).

Table 6.5: Relatives in office

Proportion of chiefs with relative holding an administrative position %

No relative in office 76%

At national Level 13%

At commune level 4%

At prefecture level 3%

Source: Village Chief Survey

Page 57: Baseline Report for the Impact Evaluation of the Central ...

57 PDCAGV IE Baseline Survey Report

Figure 6.3.2: Village chief membership in commune council

Source: Village chief survey

6.4 Village conflict Village chiefs have the important task of mediating and mitigating conflicts among villagers. They are

also at the center of administrative decision-making, and therefore play an integral part in setting the

agenda for conflict resolution mechanisms and policies. Figure 6.4.1 illustrates the prevalence of conflict

in villages as reported by village chiefs. Adultery, reported by 28% of chiefs, is the main cause of conflict.

Conflicts over theft of goods (26%) and water, theft of livestock, and debt (25% each) are also major

causes of intra-village conflict. Conflict over waste and mugging (3% each) were least reported.

Figure 6.4.1: Village conflict

Source: Village Chief Questionnaire

90%

7%

2% 1%

Figure 6.3.2: Village chief membership in commune council

Is not member Member Vice-President President

28

26

25

25

25

20

15

15

12

11

3

3

Adultery

Theft of Goods

Water

Theft of Livestock

Debt

Land

Fights between Neighbors

Employment

Inheritance

Fraud

Waste

Mugging

Figure 6.4.1: % of village chiefs reporting conflict in the last 12 months concerning...

Page 58: Baseline Report for the Impact Evaluation of the Central ...

58 PDCAGV IE Baseline Survey Report

7. Commune Mayor Survey The commune mayor questionnaire was administered to 64 community mayors (or their deputies) from

10 prefectures. It captures their attitudes, socio-economic information, and connections to their

particular villages/communes.

7.1 Demographics The average commune mayor is 50.1 years old and has been living at his/her present location for 33.2

years, though there is considerable variation around these averages.

Table 7.1.1: Commune mayor demographics

Average Standard deviation Maximum Minimum

Age of respondent 50.1 9.4 66 29

Number of years living in present location 33.2 17.8 66 0

Source: Commune mayor Questionnaire

Figures 7.1.1 and 7.1.2 illustrate the ethnic and religious affiliations of commune mayors. More than half

the commune mayors belong to either the gbaya (37%) or banda (22%) ethnic groups. A combined 41%

belong to other ethnic groups. Figure 7.1.1 compares ethnicities of commune mayors to households. It

shows that the sampled commune mayors represent the household sample fairly well.

Figure 7.1.1: Commune mayor and household ethnicity

Source: Household & commune mayor surveys

37,5

37,5

28,2

29,7

22,4

21,9

5,4

4,7

3,7

4,7

2,8

1,6

0 10 20 30 40 50 60 70 80 90 100

Households

Commune mayors

Figure 7.1.1: Commune mayors and household ethnicity (% of mayors/households)

g'baya other banda mandja g'baka other

Page 59: Baseline Report for the Impact Evaluation of the Central ...

59 PDCAGV IE Baseline Survey Report

Religious affiliation of commune mayors is shown in figure 7.1.2. More than three quarters of the

mayors are Christian, divided almost evenly between Protestants (38%) and Catholics (39%). An

additional 14% are Muslim. Figure 7.1.2 also shows that the household and mayor survey results are not

as balanced as in the case of ethnicities. For example, relatively more households are Protestant.

Figure 7.1.2: Commune mayor and household religion

Source: Household and commune mayor surveys

54,1

37,5

32,1

39,1

6,4

14,1

7,5

9,4

0 20 40 60 80 100

Households

mayors

Figure 7.1.2: Commune mayor and household religion (% of mayors/households)

Protestant Catholic Muslim Other

Page 60: Baseline Report for the Impact Evaluation of the Central ...

60 PDCAGV IE Baseline Survey Report

7.2 Education Only 14% of commune mayors did not receive any formal schooling. 28% completed primary school

while at further 45% completed secondary school (see figure 7.2.1). This is in contrast to the average

household head, who had only a 32% chance of having completed primary or higher levels of schooling

(see figure 7.2.2)

Figure 7.2.1: Education of commune mayors

Source: Household and commune mayor surveys

Figure 7.2.2: Comparison of commune mayor and household head education levels

Source: Household and commune mayor surveys

45%

28%

14%

6% 3% 2% 2%

Figure 7.2.1: Highest level of education completed by commune mayors

Secondary Primary None Higher Other Professional Religious

32,0

85,9

68,0

14,1

0 20 40 60 80 100

Household heads

Mayors

Figure 7.2.2: Education levels of commune mayors and household heads (% of mayors/household heads

reaching level)

primary education or more no primary education

Page 61: Baseline Report for the Impact Evaluation of the Central ...

61 PDCAGV IE Baseline Survey Report

79% of mayors said they could write a letter in French. 68% stated they are able to draft a letter in

another language.

Table 7.2.1: Literacy of commune mayors

Respondent can read and write letter in French 79%

Respondent can read and write letter in any other language 68%

Source: Commune mayor survey

7.3 Commune council membership and other leadership functions Of the 36% of respondents that were regular council members, 34% are presidents and 26% are vice-

presidents. Only 4% fulfill other functions (treasurer or other). Of mayors that serve on commune

councils, 92% were nominated to their position. 6% are so-called special delegates.

Figure 7.3.1: Mayors’ role in commune council

Source: Commune mayor survey

Figure 7.3.2: Mayors’ assumption of office in commune council

Source: Commune mayor survey

92%

6% 2%

Figure 7.3.2: Proportion of mayors assuming commune council positions through...

Nomination Special Delegation Other

36%

34%

26%

2% 2%

Figure 7.3.1: Mayors' role in commune council

Council Member President Vice-President Treasurer Other

Page 62: Baseline Report for the Impact Evaluation of the Central ...

62 PDCAGV IE Baseline Survey Report

Table 7.3.1 includes statistics on traditional chieftaincy and length of tenure of commune mayors. 33%

of members had been mayor since 2005, i.e. for more than 5 years. 32% of respondents also serve as a

chief or other traditional authority

As with village chiefs, commune mayors rarely hold other public positions. 52 of 64 (81%) respondents

report not to hold an additional position. 71.9% of mayors belong to the KNK party and 23.4% have no

party. 59.4% report not to have relatives in administrative positions.

Table 7.3.2: Political party membership and family involvement in public administration

Respondent is member of a political party N in %

KNK 46 71.9

Not Member 15 23.4

MDI 1 1.6

PAD 1 1.6

PSD 1 1.6

Whether someone from the respondent’s family is involved in public administration N in %

No relatives in administration 38 59.4

National level 13 20.3

Commune level 7 10.9

Military 5 7.8

Prefecture level 1 1.6

Source: Commune mayor survey

Table 7.3.1: Traditional chieftaincy and length of tenure

Respondent was mayor before 2005 33%

Whether respondent is also a traditional chief/authority 32%

Source: Commune mayor survey