Enquête Permanente Auprès des Ménages 1987-1988 (Wave 3 Panel)

Community-level data

Rainfall data associated with CILSS clusters.

Rainfall data are available for the years 1974-1988 by weather `station'. Each weather station can be linked to the CILSS Clusters. Most CILSS clusters are not located in exactly the same place as the stations with which they are associated. In such cases, the CILSS cluster is linked with the nearest station. Rainfall measurements are millimeters.

Section 0

Section 0

Section 0

Household Roster

Household Roster

Housing

Housing

Education

Education

Education

Health

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Employment

Migration

ID of Round 2 Respondents

Housing Characteristics

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Agriculture

Non-Farm Self-Employment

Non-Farm Self-Employment

Non-Farm Self-Employment

Expenditures and Inventory of Durable Goods

Expenditures and Inventory of Durable Goods

Expenditures and Inventory of Durable Goods

Expenditures and Inventory of Durable Goods

Food Expenses and Consumption of Home Production

Food Expenses and Consumption of Home Production

Fertility

Fertility

Fertility

Other Income

Other Income

Savings

Savings

Savings

Anthropometrics

Anthropometrics

ID of Panel Households

Household Expenditure Aggregates

The survey data contain all necessary information for the construction of a complete set of current accounts for each household. Since income and expenditure data are available in great detail throughout the questionnaire permitting the calculation of detailed income and expenditure aggregates, this enables, theoretically, the derivation of savings as a residual.

Given the complexity and detail involved in the different income and expenditure modules, it is possible to build household income and expenditure aggregates in different ways, each of which are legitimate but which may provide considerably different results. Thus, various researchers have constructed their own Income and Expenditure Aggregates using CILSS data. However, only one set of researchers constructed a complete set of income and expenditure aggregates for all four years of the CILSS (85-88), along with their sub-aggregate components, namely the research project "Poverty and the Social Dimensions of Structural Adjustment in Côte d'Ivoire" (RPO 675-26). Oh and Venkataraman (1992) document in detail all of those income and expenditure aggregates and sub-aggregates. The documentation includes data set names, documentation of procedures used to `clean' the data, clear the data of outliers (including information on the percentage of observations classified as outliers), and the summation procedures used to build up variables in the questionnaire into sub-aggregate level variables and finally into aggregates.

Since this set of aggregates is also the only one which is accompanied by documentation, it is the only dataset of aggregates formally available for public use. However, users should be cautioned that these data are cleared of outliers and therefore, researchers who want the presence of outliers in their data in the belief that they are meaningful, may not find this set of aggregates suitable.

Total Household Expenditure = Food Expenditure + Consumption of Home-Produced Food + Consumption of Home-Produced Non-Food Products + Other Expenditures + Paid Remittances + Wage Income in Kind

Household Income Aggregates

The survey data contain all necessary information for the construction of a complete set of current accounts for each household. Since income and expenditure data are available in great detail throughout the questionnaire permitting the calculation of detailed income and expenditure aggregates, this enables, theoretically, the derivation of savings as a residual.

Given the complexity and detail involved in the different income and expenditure modules, it is possible to build household income and expenditure aggregates in different ways, each of which are legitimate but which may provide considerably different results. Thus, various researchers have constructed their own Income and Expenditure Aggregates using CILSS data. However, only one set of researchers constructed a complete set of income and expenditure aggregates for all four years of the CILSS (85-88), along with their sub-aggregate components, namely the research project "Poverty and the Social Dimensions of Structural Adjustment in Côte d'Ivoire" (RPO 675-26). Oh and Venkataraman (1992) document in detail all of those income and expenditure aggregates and sub-aggregates. The documentation includes data set names, documentation of procedures used to `clean' the data, clear the data of outliers (including information on the percentage of observations classified as outliers), and the summation procedures used to build up variables in the questionnaire into sub-aggregate level variables and finally into aggregates.

Since this set of aggregates is also the only one which is accompanied by documentation, it is the only dataset of aggregates formally available for public use. However, users should be cautioned that these data are cleared of outliers and therefore, researchers who want the presence of outliers in their data in the belief that they are meaningful, may not find this set of aggregates suitable.

Total Household Income = Wage Income + Farm Income - Depreciation of Farm Equipment + Non-Farm Income + Non-Farm Capital Asset Depreciation + Rental Income + Income from Scholarships + Income from Remittances + Other Income.

Health Facility Data from Administrative Sources. These data contain summary statistics on 311 (out of 329) health facilities located within the 200 clusters interviewed during the 4 years of the CILSS Household Survey. The data set contains information about every health facility in the same urban "commune" as a CILSS cluster as well as about each one located within a CILSS rural cluster. Facilities near a rural cluster but not located within them, are not included. In cases where no health facility data are available for a rural cluster, information about the nearest health facility to rura l clusters can be obtained from the CILSS community data or, for 1987 only, from the health facility questionnaires completed for that year.

The data for each facility were extracted from a publication of the Direction de la Planification et des Statistiques Sanitaires in the Ministry of Public Health and Population, entitled "Annales de la Santé, 1989", as part of the research project on "The Economic and Policy Determinants of Fertility in Sub-Saharan Africa."

The Health Facility dataset includes the following information: owners hip, number of beds, number of staff of different types (doctors, paramedics, etc) and types of services offered (maternity, pharmacy, radiology, pediatrics etc). The amount of information for each facility is limited to the amount of information available from the publication. The full range of information is available for hospitals, urban hospital centers, rural hospital centers and some private facilities in Abidjan. For dispensaries and maternities, little more is offered than the name and type of facility.

The variables for the types of services available are based on staffing lists, some of which disaggregate personnel by type of service. It is possible that some smaller facilities offered a service but did not have staffing lists that were disaggregated to reflect this. Thus, a "yes" answer to a service implies that the service was definitely provided; a "no" means that the service may or may not have been provided.

PRIMARY SCHOOL INSPECTORATE (Administrative) DATA
Data collected for each primary school inspectorate linked to a CILSS cluster include the number of schools, classrooms, teachers, students and female students in the inspectorate. The data are organized by cluster and year (1985-1988). A particular cluster may have more than one primary school inspectorate associated with it, depending on the year to which the data pertain. For example, the cluster of Arrah (031) belonged to the inspectorate of Bongouanou in 1985 and 1986; but in 1987, Arrah became a new inspectorate. Therefore, cluster 031 now belongs to this new inspectorate (Arrah). Note also that if a particular cluster does not belong to a commune, then the variable NAMCOMMU (name of the commune) will be missing and NUMCOMMU (number of the commune) will be set equal to 98.

Price data

School Data from Administrative Sources. Primary school data contains information about the characteristics of primary schools nearest to or within each urban cluster; secondary school data contain similar information for secondary schools; and the inspectorate data contain information about the primary school inspectorate that covers each of the 200 clusters. These data were extracted from documents at the Côte d'Ivoire Ministry of Education and linked to each CILSS cluster as part of the data collected for the research project on "Economic and Policy Determinants of Fertility in Sub-Saharan Africa".

Data collected for the primary schools include the following information: ownership; whether the school has a library; whether there is housing available for the teachers; number of grades; number of classrooms; total number of students enrolled; and number of girls enrolled.

Only urban clusters are covered by the Primary and Secondary School Data Sets. Recall that for rural clusters, information about schools can be obtained from the CILSS Community Surveys. The Primary School Dataset contains 9055 observations and the Secondary School Dataset contains 1129 observations. The reasons for the larger than expected number of observations are primarily two-fold. First, data is gathered for up to three years (1986-88) for primary school and up to four years (1985-88) for secondary schools, per school. Secondly, all schools associated with a cluster - those located both within the cluster and nearby - are listed in the database.

School Data from Administrative Sources. Data collected for the secondary schools include the following information: ownership; whether the school has a library; whether there is housing available for the teachers; number of grades; number of classrooms; total number of students enrolled; and number of girls enrolled.

Only urban clusters are covered by the Primary and Secondary School Data Sets. Recall that for rural clusters, information about schools can be obtained from the CILSS Community Surveys. The Primary School Dataset contains 9055 observations and the Secondary School Dataset contains 1129 observations. The reasons for the larger than expected number of observations are primarily two-fold. First, data is gathered for up to three years (1986-88) for primary school and up to four years (1985-88) for secondary schools, per school. Secondly, all schools associated with a cluster - those located both within the cluster and nearby - are listed in the database.

CILSS Corrective Weights Dataset