Abstract

The Ethiopia Socioeconomic Panel Survey (ESPS) is a collaborative project between the Ethiopian Statistical Service (ESS) and the World Bank Living Standards Measurement Study-Integrated Surveys on Agriculture (LSMS-ISA) team. The objective of the LSMS-ISA is to collect multi-topic, household-level panel data with a special focus on improving agriculture statistics and generating a clearer understanding of the link between agriculture and other sectors of the economy. The project also aims to build capacity, share knowledge across countries, and improve survey methodologies and technology. ESPS is a long-term project to collect panel data. The project responds to the data needs of the country, given the dependence of a high percentage of households on agriculture activities in the country. The ESPS collects information on household agricultural activities along with other information on the households like human capital, other economic activities, and access to services and resources. The ability to follow the same households over time makes the ESPS a new and powerful tool for studying and understanding the role of agriculture in household welfare over time as it allows analyses of how households add to their human and physical capital, how education affects earnings, and the role of government policies and programs on poverty, inter alia. The ESPS is the first-panel survey to be carried out by the Ethiopian Statistical Service that links a multi-topic household questionnaire with detailed data on agriculture.

Geographic coverage

National Regional Urban and Rural

Analysis unit

• Household
• Individual
• Community

Universe

The survey covered all de jure households excluding prisons, hospitals, military barracks, and school dormitories.

Kind of data

Sample survey data [ssd]

Sampling procedure

The sampling frame for the second phase ESPS panel survey is based on the updated 2018 pre-census cartographic database of enumeration areas by the Ethiopian Statistical Service (ESS). The sample is a two-stage stratified probability sample. The ESPS EAs in rural areas are the subsample of the AgSS EA sample. That means the first stage of sampling in the rural areas entailed selecting enumeration areas (i.e., the primary sampling units) using simple random sampling (SRS) from the sample of the 2018 AgSS enumeration areas (EAs). The first stage of sampling for urban areas is selecting EAs directly from the urban frame of EAs within each region using systematic PPS. This is designed to automatically result in a proportional allocation of the urban sample by zone within each region. Following the selection of sample EAs, they are allocated by urban rural strata using power allocation which is happened to be closer to proportional allocation.

The second stage of sampling is the selection of households to be surveyed in each sampled EA using systematic random sampling. From the rural EAs, 10 agricultural households are selected as a subsample of the households selected for the AgSS, and 2 non-agricultural households are selected from the non-agriculture households list in that specific EA. The non-agriculture household selection follows the same sampling method i.e., systematic random sampling. One important issue to note in ESPS sampling is that the total number of agriculture households per EA remains at 10 even though there are less than 2 or no non-agriculture households are listed and sampled in that EA. For urban areas, a total of 15 households are selected per EA regardless of the households’ economic activity. The households are selected using systematic random sampling from the total households listed in that specific EA.

The ESPS-5 kept all the ESPS-4 samples except for those in the Tigray region and a few other places. A more detailed description of the sample design is provided in Section 3 of the Basic Information Document provided under the Related Materials tab.

Mode of data collection

Computer Assisted Personal Interview [capi]

Research instrument

The ESPS-5 survey consisted of four questionnaires (household, community, post-planting, and post-harvest questionnaires), similar to those used in previous waves but revised based on the results of those waves and on the need for new data they revealed. The following new topics are included in ESPS-5:

a. Dietary Quality: This module collected information on the household’s consumption of specified food items.

b. Food Insecurity Experience Scale (FIES): In this round the survey has implemented FIES. The scale is based on the eight food insecurity experience questions on the Food Insecurity Experience Scale | Voices of the Hungry | Food and Agriculture Organization of the United Nations (fao.org).

c. Basic Agriculture Information: This module is designed to collect minimal agriculture information from households. It is primarily for urban households. However, it was also used for a few rural households where it was not possible to implement the full agriculture module due to security reasons and administered for urban households. It asked whether they had undertaken any agricultural activity, such as crop farming and tending livestock) in the last 12 months. For crop farming, the questions were on land tenure, crop type, input use, and production. For livestock there were also questions on their size and type, livestock products, and income from sales of livestock or livestock products.

d. Climate Risk Perception: This module was intended to elicit both rural and urban households perceptions, beliefs, and attitudes about different climate-related risks. It also asked where and how households were obtaining information on climate and weather-related events.

e. Agriculture Mechanization and Video-Based Agricultural Extension: The rural area community questionnaire covered these areas rural areas. On mechanization the questions related to the penetration, availability and accessibility of agricultural machinery. Communities were also asked if they had received video-based extension services.

Cleaning operations

Final data cleaning was carried out on all data files. Only errors that could be clearly and confidently fixed by the team were corrected; errors that had no clear fix were left in the datasets. Cleaning methods for these errors are left up to the data user.

Response rate

ESPS-5 planned to interview 7,527 households from 565 enumeration areas (EAs) (Rural 316 EAs and Urban 249 EAs). However, due to the security situation in northern Ethiopia and to a lesser extent in the western part of the country, only a total of 4999 households from 438 EAs were interviewed for both the agriculture and household modules. The security situation in northern parts of Ethiopia meant that, in Tigray, ESPS-5 did not cover any of the EAs and households previously sampled. In Afar, while 275 households in 44 EAs had been covered by both the ESPS-4 agriculture and household modules, in ESPS-5 only 252 households in 22 EAs were covered by both modules. During the fifth wave, security was also a problem in both the Amhara and Oromia regions, so there was a comparable reduction in the number of households and EAs covered there.

More detailed information is available in the BID.

Abstract

The issue of food security has continual national importance in Ethiopia. To achieve social and political stability, the government has to be able to create and maintain food security by issuing an appropriate agricultural policy. Agricultural statistics is just one element that enters into this policy process to formulae, monitor, assess and evaluate the policy. The collection of reliable, comprehensive and timely data on agriculture is thus essential for the above purpose. In this regard, the Central Statistical Agency (CSA) has exerted effort to provide policy makers and users with reliable and timely agricultural data. The general objective of CSA's annual Agricultural Sample Survey (AgSS) is to collect basic quantitative information on the country's agricultural that is considered essential for development planning, socio-economic policy formulation, food security, etc. The AgSS is composed of four components: Crop production forecast survey, Main (“Mehe”) season survey. Livestock survey and survey of the “Belg” season crop area and production.

The specific objectives of the Main (“Meher”) season area and production survey are: - To estimate the total cultivated land area, production and yield per hectare of major crops (temporary). - To estimate the total farm inputs applied area and quantity of inputs applied by type for major temporary and permanent crops. - To estimate the total land used for various purposes by type of land use and the number of agricultural households, holders, members of agricultural households, average household size, average land holding per household and others

Geographic coverage

The survey covered all sedentary rural agricultural population in all regions of the country except urban and nomadic areas which were not included in the survey.

Analysis unit

Agricultural household/ Holder/ Crop

Universe

Agricultural households

Kind of data

Sample survey data [ssd]

Sampling procedure

The 1999-2000 (1992 E.C) Meher seasons annual Agricultural Sample Survey covered the rural part of the country except two zones in Afar region and six zones in Somali region that are predominantly nomadic. A two-stage stratified sample design was used to select the samples. Each zones/special wereda was adopted as stratum for which major findings of the survey are reported except the four regions; namely, Gambella, Harari, Addis Ababa and Dire Dawa which were considered as strata/reporting levels. The primary sampling units (PSUs) were enumeration area (EAs) and agricultural households were the secondary sampling units. The survey questionnaires were administered to all agricultural holders within the sample households. A fixed number of sample EAs were determined for each stratum/reporting level based on precision of major estimates and cost considerations. Within each stratum EAs were selected using probability proportional to size; size being total number of households in the EAs as obtained from the 1994 Population and Housing Census. From each sample EA, 40 agricultural households were systematically selected for the annual agricultural sample survey from a fresh list of households prepared at the beginning of the field work of the annual agricultural survey. Of the forty agricultural households, the first twenty five were used for obtaining information on area under crops. Meher and Belg season production of crops, land use, agricultural practices, crop damage, and quantity of agricultural inputs used. It is important to note that of the total forty agricultural households sampled in each of the selected EAs, data on crop cutting were collected for only the fifteen households (11th - 25th household selected).

Mode of data collection

Face-to-face [f2f]

Research instrument

The 1999-2000 annual Agricultural Sample Survey used structured questionnaires to collect agricultural information from selected sample households. List of forms in the questionnaires: - AgSS Form 92/0: Used to list all agricultural households and holders in the sample enumeration areas. - AgSS Form 92/1: Used to list selected households and agricultural holders in the sample enumeration areas. - AgSS Form 92/2: Used to collect information about crop condition. - AgSS Form 92/3A: Used to list fields and agricultural practices only pure stand temporary and permanent crops, list of fields and agricultural practices for mixed crops, other land use, quantity of improved and local seeds by type of crop and type and quantity of crop protection chemicals. - AgSS Form 92/4A: Used to collect results of area measurement. - AgSS Form 92/5: Used to list fields for selecting fields for crop cuttings and collect information about details of crop cutting. - AgSS Form 92/6: Used to collect information about cattle by sex, age and purpose

Cleaning operations

Editing, Coding and Verification: In order to insure the quality of the collected survey data an editing, coding and verification instruction manual was prepared and printed. Then 35 editors-coders and 20 verifiers were trained for two days in the editing, coding and verification operation using the aforementioned manual as a reference and teaching aid. The completed questionnaires were edited, coded and later verified on a 100% basis before the questionnaires were passed over to the data entry unit. The editing, coding and verification exercise of all questionnaires was completed in about 40 days.

Data Entry Cleaning and Tabulation: Before starting data entry, professional staff of Agricultural Statistics Department prepared edit specifications for use on personal computers utilizing the Integrated Microcomputer Processing System (IMPS) software for data consistency checking purposes. The data on the coded questionnaires were then entered into personal computers using IMPS software. The data were then checked and cleaned using the edit specifications prepared earlier for this purpose. The data entry operation involved about 35 data encoders and it took 30 days to complete the job. Finally, tabulation was done on personal computers to produce results as indicated in the tabulation plan.

Response rate

A total of 1,450 EAs (2.9% of the total EAs in the rural areas of the county) were selected for the survey. However 5 EAs were closed for various reasons beyond the control of the Authority and the survey succeeded in covering 1445 (99.7 %) EAs. With respect to ultimate sampling units, for the Meher season agricultural sample survey, it was planned to cover 36,250 agricultural households. The response rate was found to be 98.5%.

Sampling error estimates

Estimation procedures of totals, ratios and sampling errors are given in Appendix I of 1999-2000 annual Agricultural Sample Survey, Volume I report which is provided in this documentation.