Working Paper

Transcription

1 Working Paper Community-Based Monitoring System (CBMS) for Local Governance in Ghana: Results from a Case Study of Dangme West District Felix A. Asante Abena Oduro February 2006 Felix A. Asante: University of Ghana-Ghana Abena Oduro: University of Ghana-Ghana IDRC photo: N. McKee

2 Community-Based Monitoring System (CBMS) for Local Governance in Ghana: Results from a Case Study of Dangme West District By Felix A. Asante Abena Oduro February,

3 Table of Contents 1. Introduction 1.1 Background 1.2 Local Government and Decentralization System 1.3 Evaluation of Existing Poverty Monitoring System 1.4 Objectives of the Study 2. Methodology 2.1 Welfare Indicators 2.2 Design of Survey Instrument 2.3 Field Survey and Data Collection 2.4 Problems, Challenges and Lessons Learnt 3. Analysis and Results 3.1 Social and Demographic Characteristics 3.2 Education 3.3 Political Participation (Electoral Process) 3.4 Employment 3.5 Health 3.6 Reproductive Health and Child Mortality 3.7 Housing Conditions 3.8 Provision of Basic Utilities 3.9 Expenditure and Livelihood 3.10 Peace and Order 3.11 Access to Social and Community Services and Programmes 4. Conclusions Appendix 1: Community-Based Monitoring System Household Questionnaire 2

4 CHAPTER 1 Introduction 1.1 Background Since 1983, Ghana has implemented a number of programmes to stabilize the macro economy, promote growth and subsequently reduce poverty. The latest programme to be implemented by Ghana in her poverty reduction efforts is the Poverty Reduction and Growth Facility (PRGF). Ghana opted for the enhanced Highly Indebted Poor Country (HIPC) initiative of the Bretton Woods Institutions (BWIs) in February Consequently, a Poverty Reduction Strategy Paper (PRSP) was prepared and is currently being implemented. The PRSP was prepared using a consultative process. The consultation process brought to the fore specific problems based on perceptions and demands of the poor especially at the community level, which hitherto were not considered in developmental programs. One of the criteria that will be used to judge the successful implementation of the PRSP by the donors is monitoring and evaluation of programs and policies geared towards poverty alleviation. Though Ghana has benefited from a number of monitoring programs, all adopted a top-down approach to monitoring and not the bottom-up approach, i.e. from the community to the national level. This is in spite of the existence of a decentralized local government system in the country. 1.2 Local Government and Decentralization System The local government system in Ghana began in It is a three-tier local structure, in line with the decentralization policy. The first level constitutes 10 administrative regions, which is coordinated by the Regional Council. The regions are subdivided into local government assemblies District, Municipal and Metropolitan. Classification is done according to the size of the population in the area, demographic and ethnic 3

5 characteristics. The geographical areas of a municipality for instance consist of a single compact settlement. The geographical area, population and the ability of the area to provide the basic infrastructure and other development needs from internally generated monetary resources qualify the area as a metropolis. A minimum of 75,000 persons is needed for a district. There were initially one hundred and ten (110) districts, made up of metropolitans, municipalities and districts (see figure 1.1). In the last quarter of 2004, 28 new districts were created bringing the total number of districts to 138. Figure 1.1 Structure of the Local Government System in Ghana Regional Coordinating Council Metropolitan Municipal District Sub-Metropolitan District Councils Town Councils Zonal Councils Urban/Town/Area Councils Unit Committees The local government assemblies have sub-units, i.e. Zonal, Area, Town and Urban Councils. At the lowest level of the tier are about 16,000 unit committees. The unit committees are at the base structure of the local government system and represent the basic unit of planning and political administration. A unit is normally a settlement or a 4

6 group of settlements, with a population of between 500 to 1,000 in rural areas and 1,500 for the urban areas. The objective of the local government system in Ghana is to ensure that people are directly involved in the decision-making process and responsible for their own development. The District assemblies are therefore to identify community problems and development issues within their communities and to develop mechanisms for solving them. There is currently no consistent and timely data on poverty at the district and unit committee levels. Lack of data makes it difficult for the district assemblies to identify the needs of the local people and address them sufficiently. A community-based monitoring system could offer the district assemblies opportunities to assess policies they have implemented at the local level, identify problems and basic needs at the village/community level and how best they can be addressed. 1.3 Evaluation of Existing Poverty Monitoring Systems Ghana has conducted four rounds of Living Standard Surveys, which have been relied upon to assess the poverty situation in Ghana over the years. These surveys provide information on poverty trends in the country. They also provide opportunities for policy makers to trace trends in households well being over a period of time. Even though the Ghana Living Standard Surveys (GLSS) serve a purpose of providing poverty indicators, they are fraught with problems some of which are listed below. The GLSS misses out on some important poverty indicators such as voice, exclusion and malnutrition. Poor design of the questions makes them incomparable over time. Due to the high costs of implementing these surveys, their timing has been irregular and lastly; 5

7 The global nature of the Living Standard Surveys makes it gloss over poverty at the community and individual levels making it difficult for the average Ghanaian to identify him or herself with some of the results. There have been attempts to remedy some of the inadequacies of the Living Standard Surveys. For example, the Core Welfare Indicators Questionnaires (CWIQ) - which is designed to furnish policy makers with a set of simple indicators for monitoring poverty and its impact on living standards in the country help to fill in the gaps as far as some social indicators of poverty are concerned. Nevertheless, these do not still cover analysis of poverty at the community levels and at the same time the problem of regularity of data collection still persists. In addition, there have been participatory surveys, which have involved collection of data at the community level. But these have been very isolated and not on a consistent level. 1.4 Objectives of the Study Ghana has had about 18 years of decentralisation and the local government system is currently entrenched in the governance of the country. One of the main objectives of the local government system is ensure that people are directly involved in the decisionmaking process and responsible for their own development. The District assemblies have been charged to identify problems and development issues within their communities and to develop mechanisms for solving them. As is evidenced from the living standard surveys currently available in the country, very little data on community poverty exists at the district levels. The only available data on the districts are those collected by the sectoral departments for their central offices and not for the district planning offices. This major constraint in the development process makes it difficult for targeted interventions at alleviating poverty in the communities. In addition, such policy interventions use a top-down approach since it often involves very little analysis of the priorities and perceptions of the people in the communities. 6

8 Against this background, a Community-Based Poverty Monitoring System will inform policy makers, on a timely basis, of the effects of policies on the standard of living of people at the community level. This hopefully will complement the efforts of the decentralised system and achieve the main objective of local people becoming directly involved in what policies best address their needs. The proposed objectives of the CBMS-Ghana are as follows: To offer communities with simple and easy to collect poverty indicators to determine the prevailing standards of living; To offer district planning offices with up-to-date core set of welfare indicators for assessment of poverty status at the communities; To provide policy makers with data to be used for prioritisation of projects, effective planning and monitoring of developmental programmes in the various communities; Improve capacity at the district and unit committee levels in collection, processing and analysis of data collected at the local levels; To strengthen flow of information and dissemination of poverty data from the national to the committee level and finally To test a locally-feasible data processing system, without necessarily relying on central government resources. 7

9 CHAPTER 2 Methodology 2.1 Welfare Indicators Poverty in Ghana is multi-dimensional and characterized by low income, malnutrition, ill-health, illiteracy, insecurity and isolation. Most of these indicators tie in with the Minimum Basic Needs Approach identified in the literature as capturing the multidimensional characteristics of poverty. The main areas of concern were health, Water and Sanitation, Income & Livelihood, Basic Education and Literacy, Shelter, Peace and Order and Political Participation. 2.2 Design of Survey Instrument A draft household questionnaire was prepared by the CBMS-Ghana Team. This was then discussed at a workshop at the District Assembly office. Comments and suggestions from the workshop were used to improve upon the questionnaire after which it was printed for the main survey. The workshop participants were representatives from all the electoral areas in Dangme West District (see Appendix 1 for the questionnaire used for the survey). 2.3 Field Survey and Data Collection The CBMS was pilot-tested in three (3) communities in the Dangme West district; namely Dodowa, Prampram and Ningo. To create a sense of ownership and final take over of the system by the local authorities, enumerators used for the data collection were selected from the electoral areas within the communities. The District Planning Office and the Deputy District Co-ordinating Director supervised the collection of data at the local level and the CBMS-Ghana team provided training and overall supervision. 8

10 The basic sampling unit for the pilot test was the household. The collection of data was undertaken through a survey covering all households (that is, a census covering about 6,000 households) in the 3 selected communities in the Dangme West district. The CBMS survey was divided into three Phases: Phase 1 Dodowa, Phase 2 Prampram and Phase 3 Ningo. Highly intensive and interactive training sessions were conducted in each of the three selected communities by the CBMS resource team. It involved the complete discussion of a 10-page questionnaire, which seeks to gather information on a number of indicators necessary to determine prevailing poverty levels and improve the quality of life of individuals within the communities. The questions covered the following: Household Characteristics - provides information on the basic demographic characteristics on members of a household including age, gender and marital status. Education - information on levels of education and whether or not members of households are in school. Political Participation - to determine the levels of household participation and voice in the country s political processes, which encompasses both national and district level elections. Employment - types of jobs available within the communities and levels of unemployment. Health - captures the availability and accessibility of health facilities as well as common ailments prevalent in the community. Child Mortality - accessibility of mothers to postnatal and antenatal care and its effect on child mortality. Housing and Shelter - types of dwelling for households. Lighting, Water and Sanitation - access to water and sanitary facilities, which may influence the health status of households within a district. Income and Livelihood - explores the main sources of income for households and their expenditure patterns. 9

11 Peace and Order - seeks to identify main sources of conflict within the community, which may impact negatively on development. Access to Social and Community Services and Programmes - captures access to community services such as banks, telephones or post offices and programmes such as the Poverty Alleviation Fund initiated by the government to provide financial resources for small-scale entrepreneurs within the districts. The Pilot Area Dangme West District The Dangme West district is located in the southeastern part of Ghana in the Greater Accra Region. The district has a total land area of about 1,442 square kilometers. It shares boundaries with Yilo and Manya Krobo districts on the north west, Akwapim North district on the west, Tema Municipality on the south west and Dangme East district on the east. The Volta River and the Atlantic Ocean wash the northeastern and the southern portions of the district respectively. The district capital, Dodowa is about 25 kilometers from Accra, the capital of Ghana (See figure 2.1 for a map of Dangme West District). Dangme West district is one of the hottest and driest parts of the country. Temperatures are appreciably high for most parts of the year with the highest during the main dry season (November March) and the lowest during the short wet season (June August). The absolute maximum temperature is 40 degrees Celsius. Mean annual rainfall increases from millimeters on the coast to 1,220 millimeters to the north and northeast close to the foothills of the Akwapim Range. 10

12 DANGME WEST DISTRICT OF GHANA MANYA KROBO DIST. ASUOGYAMAN DIST. # Asutsuare NORTH TONGU DIST. # Lanor AKWAPIN NORTH DIST. YILO KROBO DIST. Agomeda # # Ayikuma # Doryum # Dodowa Nyigbenya # Dawa # DANGME EAST DIST. GGDGDGDGD N TEMA DIST. # Afienya Dawhenya # Prampram # Omankope # Old Ningo # G U L F O F G U I N E A # Major town Road District ff 11

13 The unreliability and dependence of farmers on rain water makes farming a vulnerable occupation. Periodic main crop failures are common phenomena even in the betterwatered northern parts. The predominant vegetation type found in the district is of the short grass savannah interspersed with shrubs and short trees, a characteristic of the subsahelian type. The soils are highly elastic when wet but become hard and compact when dry and then crack vertically from the surface. This renders the soil unsuitable for hand cultivation. The main occupational activity of the economically active population is agriculture (crop farming, livestock and fishing). The total population of Dangme West district is 98,809 (2000 Population and Housing Census). Generally, the district has a lower population density than the average for the country, 55.3 persons per square kilometer against the national average of 63 persons per kilometer. Of the total population in 2000, 48.2 percent are males and 51.8 percent females. The dependency ratio (proportion of the population of ages 0-14 and 65+ years to the economically active population, years) is The Dangme West district is more rural than urban. According to the 2000 population census, 76 percent of the population live in rural areas whiles 23.6 percent live in the urban areas. 2.4 Problems, Challenges and Lessons Learnt Selection of Enumerators the most important challenge is the selection of enumerators within the district. Assistance from the Assembly in the Dangme West district proved invaluable since the CBMS team had no knowledge of what local capacity prevailed. The enumerators identified by representatives of the electoral areas had low educational levels and were inexperienced so teachers within the electoral areas were used to administer the questionnaire. The team focused on training local teachers as enumerators for two reasons. First, they are literate both in English and the local language and can therefore translate the questionnaire into the local dialect to households. Secondly, they are often wellknown and respected in the various communities. Although the choice of teachers as enumerators is laudable, the main challenge for them arises when schools are in 12

14 session and they have to juggle between the two responsibilities. Thus, the available times for administering the questionnaires are either after school or during the weekends. Community Demarcation Community demarcation within the districts was not distinct in some of the areas surveyed. This created situations where the CBMS team either overestimated or underestimated the number of enumerators required for the fieldwork. This occurred in Dodowa and Ningo. Multiplicity of Surveys Enumerators indicated that households are often inundated with various kinds of surveys, which do not result in the provision of or improvement in services within the communities. For instance, during the CBMS poverty survey within the Old Ningo community, another survey on district health insurance was also in progress. Data Verification and Entry there was no local capacity for data entry in all the traditional areas where the surveys were conducted. This placed enormous pressure on the team and required that at the end of the field work, trainers had to go through the filled questionnaires with each enumerator to ensure that the questionnaires were filled correctly. Data was then transported to Accra for collation. It will be useful to provide minimum training in data collation to a select team within the districts before districts can successfully engage in their own poverty monitoring. Compensation enumerators demanded higher compensation for administering the questionnaires because they claimed they had to travel long distances to visit households and also had to visit households more than once in order to get them completed. 13

15 CHAPTER 3 Analysis and Results 3.1: Social and Demographic Characteristics The total number of households surveyed in the three communities was The female population surveyed exceeded the male population in all three communities. In Dodowa, out of 8,409 people in the sample population 52.7% were female. The number of females in the sample population of Ningo was 11.3 percent more than the males. Of the 6,649 people sampled in Prampram 55.2% were female and 44.8% male (Table 3.1). Even though the population had more females in all the three communities, males dominated as household heads in Dodowa and Prampram, 61.2% and 56.4%, respectively. In Ningo, females dominated as household heads, 53.9%. Ningo had the highest mean household size of 4.12 while the lowest mean household size of 3.48 was recorded from Dodowa. Generally, 45% of the population had never married in all 3 communities but in Prampram approximately half the population sampled had never married. A third of the sampled population was married with about 5% widowed. In Dodowa and Ningo, more than half the population sampled had married or been in a form of union. In Prampram, however, less than half the population had been in any type of union. The highest rate of divorcees, 7% were in Dodowa with Prampram recording the lowest of 3.6%. In Dodowa, the majority of the population aged 15 years and above were married which was about 15% more than the proportion that was never married. The percentage of the population in informal/loose union was 4.7% and the proportion that was divorced/separated marital status, 9.4%. The distribution of the marital status of the population 15 years and above in Ningo and Prampram was similar to that of Dodowa, except for informal/loose union and divorced/separated cases where there were equal 14

16 percentages. While Prampram had the highest percentage of married population (53.1%), Ningo had the highest widowed (8.1%) and Dodowa the highest percentage of never married (Table 3.1). Table 3.1: Social and Demographic Characteristics of Sample Indicator Unit Dodowa Ningo Prampram Household/Demographic Characteristics No. of Households Sample Population Male Female No. No. % % 2,415 8, ,996 8, ,319 8, Head of Household Male Female % % Average Household Size Marital Status (15 years & above) Married Informal/Loose Union Divorced/Separated Never Married Widowed % % % % % : Education In all the three communities, more than half the sample said they were literate, that is they could read and/or write. The highest literacy rate was in Dodowa and the lowest in Ningo (Table 3.2). Of the 34.8% of the sample in Dodowa who could not read and/or write, 61.9% were female and 38.1% male. In Ningo, there were 26.2% more illiterate females than males. Prampram also recorded about 19% more female illiterates than males. 15

17 The level of school attendees was generally higher than the literacy rate by 12.2% in Dodowa, 12.2% in Ningo and 13.6% in Prampram. Of those who had never attended school, the females were higher than males in the three communities by about 24% in Prampram, 31.2% in Ningo and 28.2% in Dodowa. Generally, literacy and school attendance were lowest in Ningo with Dodowa being the most literate and having a high school attendance rate. Table 3.2: Educational Status of Households (%) Literacy (read and/or write) Yes Dodowa Ningo Prampram Male Female No Male Female Ever Attended School Yes Male Female No Male Female Net Enrolment Rate (Basic Education) 1 Male Female Defined as the population aged 6 to 14 years old currently in school (i.e. primary 1 to JSS 3) divided by the population aged 6 to 14 years in the community. 16

18 The net enrolment rate for basic education, 6 years to 14 years for males and females were generally close in all the three communities. Whereas the net enrolment rate of boys in Dodowa and Ningo was greater than that for girls, in Prampram, it was the reverse with the net enrolment rate of girls being higher than that of boys (Table 3.2). The most frequently provided reason for not attending school was affordability, i.e. 33.1% in Dodowa, 42.2% in Ningo and 37.5% in Prampram (Table 3.3). The decision of parents not to send child to school was about a third of the reasons in all the three communities. About 10% of the school going population in Prampram was not interested in attending school. Table 3.3: Reasons for Not Ever Attending School (%) Reasons Dodowa Ningo Prampram Parents deliberately refuse to send child to school Cannot afford No importance of Education Not interested School too far Others : Political Participation (Electoral Process) Generally, the majority of respondents participated in the electoral process (Table 3.4). Almost 32% and 14% more of the population sampled participated in Prampram than in Dodowa and in Ningo respectively. Females dominated in the electoral process in all the three communities surveyed. In Dodowa, the population of males as against females who did not participate was equal. This contrasts with Ningo where 17% more females did not participate in the electoral process than males. About 11% more females did not 17

19 participate in Prampram. Female participation in the electoral process was highest in Prampram and lowest in Dodowa. Even though the participation was highest in Prampram, the percentage of females relative to males was the lowest. Out of the 11% who did not participate in the electoral process, about 44% had other personal reasons. In Ningo, about a fifth of the population who did not participate in the electoral process were seasonal migrants and 3% could not authenticate their identification. In Dodowa, 16% of the population who did not participate in the process had no reason. Identification difficulties made participation impossible for about 12% and 9% had no interest in participating. Table 3.4: Participation in Electoral Process (%) Participation/Reason Dodowa Ningo Prampram Participation Yes Male Female No Male Female Reasons for not participating No reason Not interested Sick on admission Identification in doubt Seasonal migrant Other : Employment Wholesale/Retail trade and agriculture (including forestry and fishing) were the most important employment activities in the three communities surveyed (Table 3.5). 18

20 Agriculture was highest in Ningo (53%) and lowest in Prampram (28%). In Dodowa about 30% were employed in agriculture. Trading activities had almost the same percentage of the population in all the three communities. Dodowa had about 29% of the population followed by Prampram with 27% and then Ningo with 22%. Community/Social service was the third most important employment activity in Dodowa, Ningo and Prampram. In Prampram 26% of the population was employed in this activity compared to 11% in Dodowa and 7% in Ningo. Disaggregating the type of employment activity by gender shows that generally males dominated in agriculture/forestry/fishing and construction while females dominated in wholesale/retail trade and fish processing Table 3.5: Type of Employment Activity (Industry) (Percent) Activity Agric/Forestry/Fishing Mining & Quarrying Manufacturing Construction Transport/Storage Wholesale/Retail Trade Finance/Insurance/Services Electricity, Gas & Water Community/Social Service Fish Processor Dodowa Male Female Total Ningo Male Female Total Prampram Male Female Total : Health Just over a quarter of the population in Dodowa and Ningo reported being ill in the four weeks preceding the survey (Table 3.6). In both towns the proportion of women who reported being unwell was higher than the proportion of men. Compared to the other two towns, in Prampram a significantly lower proportion of the population reported illness, with a higher proportion of women reporting illness compared to men. 19

21 The most frequently reported cause of illness was fever/malaria. The proportion that reported this ranged between 70% in Prampram and 63% in Ningo. Gastro-intestinal infection was the second most frequently reported condition. There was no significant difference between women and men in the pattern of illness. Table 3.6 Incidence and Cause of Ill-health (%) Dodowa Ningo Prampram Been Ill in last four weeks Male Female All Type of Sickness Fever/Malaria Men Women Gastro-intestinal Men Women Coughing Skin Condition Injury/Accident Health Provider Consulted Hospital Men Women Health Centre Doctor Pharmacist Spiritual/Traditional Healer In both Ningo and Prampram the majority of reported cases of illness were attended to at a hospital (private, public or mission). With the exception of Ningo there was no significant difference between the proportion of men and women that attended hospital when ill. Second in importance as a place to seek medical attention when ill was the pharmacist or drugstore. Third in importance was the health centre. 3.6: Reproductive Health and Child Mortality The proportion of women who have ever been pregnant increases with age (Figure 3.1). After the age of 40 years, there appears to be no significant difference amongst the age groups. Women in Ningo are more likely to have ever been pregnant compared to women 20

22 in Dodowa and Prampram. When women are grouped into five year age-groups, the likelihood of a woman within an age-group being pregnant in the last 12 months increases and peaks for the year age group in Ningo (Figure 3.2). Amongst women in Prampram, the incidence of pregnancy in the last 12 months peaks within the age-group. In Dodowa, women within the and age groups had the highest incidence of pregnancy in the last 12 months (Figure 3.2). Figure 3.1: Proportion of women within Age-Group who have ever been Pregnant Proportion of Women within Age-Group who have ever been Pregnant Percentage Dodowa Ningo Prampram years years years years years years years 50 years and above Age Figure 3.2: Proportion of Women within Age-Group who were Pregnant in the last 12 months Proportion of Women within Age-Group who were Pregnant in last 12 months Percentage Dodowa Ninigo Prampram 5 0 Below 15 years years years years years years years years Age 21

23 Figure 3.3: Proportion of Women within Age-Group that had Live Births in last 12 months Proportion of women within age-group that had live births Percent years years years years years years years Dodowa Ningo Prampram Age Group The proportion of women who had live births in the 12-month period peaked amongst the age-group in Dodowa and Ningo and peaked amongst the age group for women in Prampram (Figure 3.3). In Dodowa and Prampram, approximately three-quarters of the births were delivered in either a maternity home or health centre (Table 3.7). The proportion was significantly lower in Ningo. In contrast to the other two communities, a significantly larger proportion of the women in Ningo gave birth to their babies with the assistance of a traditional birth attendant or at home. Almost all expectant mothers received ante-natal care. Attendance at post-natal clinic was relatively lower. The incidence of reported cases of maternal mortality by the households was lowest in Prampram. Less than 1% of households in Prampram reported losing a mother during childbirth compared to 1.5% and 1.6% respectively in Dodowa and Ningo (Table 3.7). 22

24 Table 3.7: Indicators of Reproductive Health (%) Dodowa Ningo Prampram Where birth of child took place Traditional Birth Attendant Maternity Home Health Centre At home Other Received Ante-natal care Received Post-natal care Maternal Mortality % of Households that had lost a mother during child birth Households were asked to indicate the age at which children died and to indicate if any were lost before birth. A significant proportion of pregnancies are lost before full term. In Dodowa, Ningo and Prampram, 23.6%, 17.8% and 16.2% respectively of reported child mortalities occurred before the birth of the child. Of children that survived childbirth, more than half of the mortalities in Dodowa and Ningo occurred before the first birthday. In Prampram, the proportion was about 46%. The mortality rate for infants a month old is not insignificant particularly in Dodowa and Prampram (Figure 3.4). Figure 3.4: Age-Profile of Child Mortality Age-Profie of Child Mortality Proportion of Children Dodow a Ningo Prampram month 2 months 3 months 4 months to 1 year Age of Child at Death Over 1 year to 5 years 23

25 3.7: Housing Conditions The majority of households in all three towns reside in compound houses. A larger proportion of households headed by women reside in compound houses compared to households headed by men. In the same instance a smaller proportion of households headed by women reside in either detached/semi-detached houses or apartments (Table 3.8). Not counting bathrooms, toilets and kitchens, more than half of the households in the three towns reside in single rooms. Households headed by women are more likely to have one room in addition to the utility rooms. The mean number of persons per room tends to be lower for households headed by women, suggesting that this category of households has a lower mean size. Most households in the three towns reside either in owner occupied or rental accommodation. In all three communities, a significantly higher proportion of households headed by women live in property they own compared to households headed by men. Households headed by women are also more likely to be living in accommodation that is rent free compared to households headed by men (Table 3.8). In Dodowa almost all households occupy houses that are roofed with iron sheets. In Prampram and Ningo roofing is either with iron sheets or asbestos. More than half the households in these two communities live in houses that are roofed with asbestos roofing sheets (Table 3.8). Dodowa stands out amongst the three as having the largest proportion of households living in houses that have mud walls. This contrasts with Ningo and Prampram where houses made of mud walls are the exception (Table 3.8). 24

26 Table 3.8: Housing Conditions Dodowa Ningo Prampram Male Female All Male Female All Male Female All Type of Dwelling of Household Detached/Semi-detached House or Apartment Compound House Huts/Building in same compound Huts/Buildings in different compounds Other Number of Rooms per Household No room Single rooms rooms rooms or more rooms Occupancy Status Own house Renting Provided rent free Perching Other Material of the Roof of the House Mud Thatch Wood Iron Sheets Cement/Concrete Roofing tiles Asbestos Other Material of the Walls of the House Mud/Mud bricks Stone Burnt bricks Cement/Sandcrete Wood/Bamboo Iron Sheets Cardboard Other : Provision of Basic Utilities Electricity is the main source of lighting for most households, although a significantly lower proportion of households headed by women use electricity for lighting. Next in importance as a source of lighting is kerosene/gas (Table 3.9). Wood-based products are the main source of fuel for cooking in at least 80% of the households in the three towns (Table 3.9). The majority of households obtain drinking water from either a pipe in the dwelling or an outdoor tap. Almost 90% of the households in Ningo and over 98% in Prampram obtain 25

27 their drinking water from these sources. In Dodowa, there is relatively greater reliance on boreholes and protected wells compared to the other towns. None of the households in Ningo reported drinking water from rivers/lakes and ponds. About 1% of households in Dodowa and 0.4% in Prampram however made recourse to these sources to obtain water for drinking. Table 3.9: Basic Utilities available to Households (%) Dodowa Ningo Prampram Male Female All Male Female Male Female All Main Source of Lighting Electricity Generator Kerosene/Gas Lantern Candles/torches Biogas Osono Other Main fuel for cooking Wood Charcoal Gas Electricity Kerosene Other Main source of drinking water Piped into dwelling Public outdoor tap Borehole Protected well Unprotected well River/lake/pond Vendor/truck Other Method of refuse disposal Collected Dumped by household Burned by Household Buried by Household Other Type of toilet used by facilities Flush Toilet Covered pit latrine Uncovered pit latrine Pan/bucket KVIP No toilet (bush/beach) Other Access to what may be described as safe sanitation is low in Prampram and Ningo. A minority of households in these towns use flush toilets, covered pit latrines or the KVIP (Table 3.9). Households dispose of refuse by dumping. The environmental implication of this is particularly alarming if the dumped refuse is not collected regularly or the dumping site is near a water source. 26

28 3.9: Expenditure and Livelihood On the average, monthly household expenditures falls between 1,073,193 and 1,382,319 in Dodowa, Ningo and Prampram (Table 3.10). Prampram has the highest mean monthly expenditures of 1,382,319 followed by Ningo and Dodowa, respectively. In Dodowa, mean monthly food expenditures (actual) form 43.0% of total expenditures while imputed food expenditures is 5.20%. Thus food is the most important item in the household expenditures followed by non-food expenditures with 46.3%. This expenditure pattern is present in Ningo and Prampram. In Ningo, 47.4% of the mean household expenditure is on food (actual) and 9.64% is on imputed food. With Prampram, 51.5% is spent on food (actual) and 3.79% is spent on imputed food. Non-food items form 38.0% in Ningo and 41.5% in Prampram. The share of remittances in total mean monthly household expenditures was the highest in Dodowa, 2.83% followed by Ningo, 1.83% and lastly Prampram with 1.16%. The expenditure pattern reflects the type of employment activities the household engaged in. For example, Ningo has the highest percentage of its population in agriculture/forestry/fishing (53.1%) and thus the highest imputed food expenditures (9.64%) in the three communities surveyed. Table 3.10: Mean Monthly Household Expenditures (cedis) and Percentage Shares Expenditure Items Dodowa Ningo Prampram Food (Actual) 461, , ,

29 (43.00) (47.42) (51.53) Food (Imputed) 56, (5.20) 129, (9.64) 52, (3.79) Housing 28, (2.64) 41, (3.08) 26, (1.93) Non-Food Education Health Water Lighting Garbage disposal Toilet Facility Transport Others 204, , , , , , , , (46.33) 149, , , , , , , , (38.03) 174, , , , , , , , (41.59) Remittances 30, (2.83) Total 1,073, (100) 24, (1.83) 1,345, (100) 16, (1.16) 1,382,319 (100) Table 3.11: Mean Monthly Household Expenditure (cedis) by Sex`of Household Head Gender Male Mean Monthly Expenditure Dodowa Ningo Prampram 1,259, ,340, ,474,379.7 Female 779, ,350, ,263,423.3 Total 1,073, ,345, ,382,319.4 Disaggregating the mean monthly household expenditure by the sex of the household head shows that the mean monthly expenditure for male-headed households is higher in Dodowa and Prampram than for households headed by women. In Ningo, the mean 28

30 monthly household expenditure is slightly higher for female-headed households than for male-headed (Table 3.11). 3.10: Peace and Order Almost 7% of households in Dodowa and Prampram reported that at least one member had been a victim of violent crime. Amongst households in Ningo the proportion was 4.3%. The most frequently reported crime in Dodowa and Ningo was robbery. In Prampram, assault was the most frequently reported crime, i.e. 73%. Land disputes were identified as the most frequent causes of conflict in the three towns. Marriage is the second most frequently reported cause of dispute in Ningo and Prampram; whilst indebtedness is the second most frequently reported cause in Dodowa (Table 3.12). When there is conflict most of the cases are reported to the police station. This is particularly the case in Dodowa. Next in importance as a recourse for help is an elderly person in the community. Thus both modern and informal conflict resolution mechanisms are utilized (Table 3.12). 3.11: Access to Social and Community Services and Programmes In Dodowa and Prampram, about 40% of households have members who own a bank account. In Ningo however, less than a third of the households have members with a bank account (Table 3.13). Susu membership is lower than ownership of a bank account in the three towns. In Dodowa, 27% of households that own a bank account are members of susu, and 53% of households that have members who belong to a susu group own bank accounts. In Ningo and Prampram, the proportion of households with susu members who own bank accounts is 48%. Ownership of a post office address is quite low in the three towns. However, almost 40% of households in Prampram have access to public telephones. The proportion though is significantly lower in Dodowa and Ningo. A very small proportion of the population have 29

31 telephones installed in their homes. The ownership of mobile phones is much larger at approximately 16%. Table 3.12: Peace and Conflict (%) D odowa N ingo Pram pram P roportion of households that had m em bers w ho are victim s of crim e T ype of C rim e R obbery D efraud A ssault O ther M ajor cause of conflict Indebtedness E thnic C onflict P olitical D ifferences M arriage Land dispute C hieftaincy R eligion Fishing dispute T heft D estruction of farm by livestock O ther M ain source of help w hen conflict occurs Com m unity/village authorities D istrict Authorities P olice Station R elatives E lderly P erson in Com m unity Chief Fisherm an O ther M ajor cause of conflict (num ber) Indebtedness E thnic C onflict P olitical D ifferences M arriage Land dispute C hieftaincy R eligion Fishing dispute T heft D estruction of farm by livestock O ther M ain source of help w hen conflict occurs Com m unity/village authorities D istrict Authorities P olice Station R elatives E lderly P erson in C om m unity Chief Fisherm an O ther A minority of households have members that have benefited from projects of the Social Investment Fund, Village Infrastructure Project or the Poverty Alleviation Fund. The first two are donor funded projects that operate in every district. The low proportion of self- 30

32 identified beneficiaries may be due partly to insufficient information about the project. Social programmes sponsored by the district assembly or local community or NGOs benefit a limited number of households (Table 3.13). Coverage of health insurance schemes is wider ranging from 22.5% in Prampram to 29.3% in Dodowa and 31.5% in Ningo. Most households are aware of the national health insurance scheme but are not willing to register for a number of reasons. Lack of interest in the scheme and not having enough money to register are the most frequently given reasons for not registering with the scheme (Table 3.13). Table 3.13: Access to Social and Community Services and Programmes (%) D o d o w a N in g o P r a m p ra m M e m b e r s o f th e H o u s e h o ld h a v e th e fo llo w in g : B a n k A c c o u n t P o s t o ffic e a d d r e s s M o b ile T e le p h o n e L a n d -lin e a t h o m e P u b lic p h o n e S u s u s M e m b e r o f h o u s e h o ld h a s b e n e fite d fro m th e ff: S o c ia l In v e s tm e n t F u n d V illa g e In fra s tr u c tu re P ro je c t P o v e rty A lle v ia tio n F u n d S o c ia l P ro g ra m m e o f th e D is tric t/lo c a l c o m m u n ity S o c ia l p r o g r a m m e o rg a n is e d b y N G O P ro p o rtio n o f h o u s e h o ld s w ith N H IS m e m b e rs R e a s o n s fo r n o t re g is te rin g w ith H e a lth In s u ra n c e S c h e m e N o t re a d y N o t in te r e s te d D o n o t k n o w a b o u t th e s c h e m e N o t a ro u n d w h e n r e g is tra tio n to o k p la c e D o n t h a v e m o n e y to re g is te r B e n e fit fro m th e S S N IT O th e r

33 CHAPTER 4 Conclusions The census of households in Dodowa, Ningo and Prampram covered about a quarter of the population in the Dangme West District. The census has revealed similarities and diversities amongst the population in the three towns. In the three towns, the literacy rate amongst women tended to be lower than men although in Dodowa the proportion of literates was highest. The proportion of parents who would not send their children to school was about the same across the three towns. Women were more likely to vote in all the three towns, although Dodowa had a significantly lower voter participation rate. Fever/malaria was the most frequently reported illness in the three towns. Although the incidence of flush toilets was broadly the same in the three towns, the residents in Dodowa had greater access to safe sanitation than did the residents of the other towns. The extent of variation within the district reveals the importance of spatially disaggregated data. Dangme West is located in the region that had the lowest headcount poverty index in 1998/99, however some of the poverty indicators in the district for example access to safe sanitation) are not significantly different from indicators in poorer regions and districts. Effective planning must be informed by relevant and up-to-date data and information. Spatially disaggregated is critical to ensuring that deprived households and locations within the district are targeted. This pilot programme has provided useful baseline data for poverty monitoring and evaluation. The value of this data base will be increased further if such an exercise is conducted at relevant intervals, possibly every two or three years. 32

34 APPENDIX 33

35 CEPA A - GENERAL INFORMATION QUESTIONNAIRE REF. NUMBER ENUMERATOR S NAME: RESPONDENT S NAME AND ADDRESS: REGION: DISTRICT: ENUMERATION AREA (EA): TOWN/VILLAGE: HOW LONG HAS THE HOUSEHOLD BEEN IN SINCE (YEAR) THE TOWN/VILLAGE OR EA DATE OF INTERVIEW: TIME OF INTERVIEW: FROM TO: ANY COMMENTS: