Abstract

Survey based Harmonized Indicators (SHIP) files are harmonized data files from household surveys that are conducted by countries in Africa. To ensure the quality and transparency of the data, it is critical to document the procedures of compiling consumption aggregation and other indicators so that the results can be duplicated with ease. This process enables consistency and continuity that make temporal and cross-country comparisons consistent and more reliable.

Four harmonized data files are prepared for each survey to generate a set of harmonized variables that have the same variable names. Invariably, in each survey, questions are asked in a slightly different way, which poses challenges on consistent definition of harmonized variables. The harmonized household survey data present the best available variables with harmonized definitions, but not identical variables. The four harmonized data files are

a) Individual level file (Labor force indicators in a separate file): This file has information on basic characteristics of individuals such as age and sex, literacy, education, health, anthropometry and child survival. b) Labor force file: This file has information on labor force including employment/unemployment, earnings, sectors of employment, etc. c) Household level file: This file has information on household expenditure, household head characteristics (age and sex, level of education, employment), housing amenities, assets, and access to infrastructure and services. d) Household Expenditure file: This file has consumption/expenditure aggregates by consumption groups according to Purpose (COICOP) of Household Consumption of the UN.

Geographic coverage

National

Analysis unit

• Individual level for datasets with suffix _I and _L
• Household level for datasets with suffix _H and _E

Universe

The survey covered all de jure household members (usual residents).

Kind of data

Sample survey data [ssd]

Sampling procedure

Sample Frame The list of households obtained from the 2001/2 Ethiopian Agricultural Sample Enumeration (EASE) was used as a frame to select EAs from the rural part of the country. On the other hand, the list consisting of households by EA, which was obtained from the 2004 Ethiopian Urban Economic Establishment Census, (EUEEC), was used as a frame in order to select sample enumeration areas for the urban HICE survey. A fresh list of households from each urban and rural EA was prepared at the beginning of the survey period. This list was, thus, used as a frame in order to select households from sample EAs.

Sample Design For the purpose of the survey the country was divided into three broad categories. That is; rural, major urban center and other urban center categories.

Category I: Rural: - This category consists of the rural areas of eight regional states and two administrative councils (Addis Ababa and Dire Dawa) of the country, except Gambella region. Each region was considered to be a domain (Reporting Level) for which major findings of the survey are reported. This category comprises 10 reporting levels. A stratified two-stage cluster sample design was used to select samples in which the primary sampling units (PSUs) were EAs. Twelve households per sample EA were selected as a Second Stage Sampling Unit (SSU) to which the survey questionnaire were administered.

Category II:- Major urban centers:- In this category all regional capitals (except Gambella region) and four additional urban centers having higher population sizes as compared to other urban centers were included. Each urban center in this category was considered as a reporting level. However, each sub-city of Addis Ababa was considered to be a domain (reporting levels). Since there is a high variation in the standards of living of the residents of these urban centers (that may have a significant impact on the final results of the survey), each urban center was further stratified into the following three sub-strata. Sub-stratum 1:- Households having a relatively high standards of living Sub-stratum 2:- Households having a relatively medium standards of living and Sub-stratum 3:- Households having a relatively low standards of living. The category has a total of 14 reporting levels. A stratified two-stage cluster sample design was also adopted in this instance. The primary sampling units were EAs of each urban center. Allocation of sample EAs of a reporting level among the above mentioned strata were accomplished in proportion to the number of EAs each stratum consists of. Sixteen households from each sample EA were inally selected as a Secondary Sampling Unit (SSU).

Category III: - Other urban centers: - Urban centers in the country other than those under category II were grouped into this category. Excluding Gambella region a domain of "other urban centers" is formed for each region. Consequently, 7 reporting levels were formed in this category. Harari, Addis Ababa and Dire Dawa do not have urban centers other than that grouped in category II. Hence, no domain was formed for these regions under this category. Unlike the above two categories a stratified three-stage cluster sample design was adopted to select samples from this category. The primary sampling units were urban centers and the second stage sampling units were EAs. Sixteen households from each EA were lastly selected at the third stage and the survey questionnaires administered for all of them.

Mode of data collection

Face-to-face [f2f]

Abstract

The HICE survey basically reflect the income dimension of poverty while WM survey aims at providing socioeconomic data that reflect the non-income dimension of poverty. The HICE survey provides statistics on income, consumption and expenditure of households and WM survey provides basic indicators on the various socioeconomic areas including health, education, nutrition, access to and utilization and satisfaction of basic facilities/services and related non-income aspects of poverty. The HICE survey has been conducted together with the WM survey every four-five years since 1995/96. The latest of these HICE surveys is for 2004/5 and covered a representative sample of 21,600 households. Previous HICE were similarly representative, covered 11,928 and 17,332 households for 1995/96 and 1999/00, respectively.

Unlike the previous two HICE surveys that had been conducted in 1995/96 and 1999/00, in the 2004/05 HICE survey data on Household Consumption Expenditure and Household Income were collected independently using separate modules. However, this statistical report concentrates only on the household consumption expenditure part.

The core objective of the HICE survey is to provide data that enable to understand the income aspects of poverty and the major objectives are to: - assess the level, extent and distribution of income dimension of poverty; - provide data on the levels, distribution and pattern of household expenditure that will be used for analysis of changes in the households' living standard level over time in various socio-economic groups and geographical areas; - provide basic data that enables to design, monitor and evaluate the impact of socio- economic policies and programs on households/individuals living standard; - furnish series of data for assessing poverty situations, in general, and food security, in particular; - provide data for compiling household accounts in the system of national accounts, especially in the estimation of private consumption expenditure; and - obtain weights and other useful information for the construction of consumer price indices at various levels and geographical areas.

Geographic coverage

The 2004/05 HICE Sample Survey covered all rural and urban parts of the country except all zones of Gambella Region, and the non-sedentary population of three zones of Afar and six zones of Somali regions.

Analysis unit

• Households
• Individuals
• Consumption expenditure commodities/ services

Universe

The survey covered all households in the selected sample areas excluding residents of collective quarters, homeless persons and foreigners.

Kind of data

Sample survey data [ssd]

Sampling procedure

The list of households obtained from the 2001/2 Ethiopian Agricultural Sample Enumeration (EASE) was used as a frame to select EAs from the rural part of the country. On the other hand, the list consisting of households by EA, which was obtained from the 2004 Ethiopian Urban Economic Establishment Census, (EUEEC), was used as a frame in order to select sample enumeration areas for the urban HICE survey. A fresh list of households from each urban and rural EA was prepared at the beginning of the survey period. This list was, thus, used as a frame in order to select households from sample EAs.

Sample Design For the purpose of the survey the country was divided into three broad categories. That is; rural, major urban center and other urban center gories.

Category I: Rural: - This category consists of the rural areas of eight regional states and two administrative councils (Addis Ababa and Dire Dawa) of the country, except Gambella region. Each region was considered to be a domain (Reporting Level) for which major findings of the survey are reported. This category comprises 10 reporting levels. A stratified two-stage cluster sample design was used to select samples in which the primary sampling units (PSUs) were EAs. Twelve households per sample EA were selected as a Second Stage Sampling Unit (SSU) to which the survey questionnaire were administered.

Category II:- Major urban centers:- In this category all regional capitals (except Gambella region) and four additional urban centers having higher population sizes as compared to other urban centers were included. Each urban center in this category was considered as a reporting level. However, each sub-city of Addis Ababa was considered to be a domain (reporting levels). Since there is a high variation in the standards of living of the residents of these urban centers (that may have a significant impact on the final results of the survey), each urban center was further stratified into the following three sub-strata.

Sub-stratum 1:- Households having a relatively high standards of living Sub-stratum 2:- Households having a relatively medium standards of living and Sub-stratum 3:- Households having a relatively low standards of living.

The category has a total of 14 reporting levels. A stratified two-stage cluster sample design was also adopted in this instance. The primary sampling units were EAs of each urban center. Allocation of sample EAs of a reporting level among the above mentioned strata were accomplished in proportion to the number of EAs each stratum consists of. Sixteen households from each sample EA were finally selected as a Secondary Sampling Unit (SSU).

Category III: - Other urban centers: - Urban centers in the country other than those under category II were grouped into this category. Excluding Gambella region a domain of “other urban centers” is formed for each region. Consequently, 7 reporting levels were formed in this category. Harari, Addis Ababa and Dire Dawa do not have urban centers other than that grouped in category II. Hence, no domain was formed for these regions under this category.

Unlike the above two categories a stratified three-stage cluster sample design was adopted to select samples from this category. The primary sampling units were urban centers and the second stage sampling units were EAs. Sixteen households from each EA were lastly selected at the third stage and the survey questionnaires administered for all of them.

Sample Size and Selection Scheme Category I: - Totally 797 EAs and 9,564 households were selected from this category. Sample EAs of each reporting level were selected using Probability Proportional to Size (PPS) with systematic sampling technique; size being number of households obtained from the 2001/2 Ethiopian Agricultural Sample Enumeration. From the fresh list of households prepared at the beginning of the survey 12 households per EA were systematically selected and surveyed.

Category II: - In this category 485 EAs and 7,760 households were selected. Sample EAs from each reporting level in this category were also selected using probability proportional to size with systematic sampling method; size being number of households obtained from the 2004 EUEEC. From the fresh list of households prepared at the beginning of the survey 16 households per EA were systematically selected and covered by the survey.

Category III:-127 urban centers, 275 EAs and 4,400 households were selected in this category. Urban centers from each domain and EAs from each urban center were selected using probability proportional to size with systematic sampling method; size being number of households obtained from the 2004 EUEEC. From the listing of each EA 16 households were systematically selected and the survey was carried out on the 16 ultimately selected households.

Including region rural, region urban and country domains, totally 61 reporting levels (including the 10 sub-cities of Addis Ababa) were formed. For the overall distribution of planned and covered EAs and households see Annex I of the 2004-2005 Household Income, Consumption and Expenditure Survey (HICE).

Mode of data collection

Face-to-face [f2f]

Research instrument

The Household Income, Consumption and Expenditure Survey questionnaire contains the following forms: - Form 1: Area Identification and Household Characteristics - Form 2A: Quantity and value of weekly consumption of food and drinks consumed at home and tobacco/including quantity purchased, own produced, obtained, etc for first and second week.
- Form 2B: Quantity and value of weekly consumption of food and drinks consumed at home and tobacco/including quantity purchased, own produced, obtained, etc for third and fourth week .
- Form 3A: All transaction (income, expenditure and consumption) for the first and second weeks except what is collected in Forms 2A and 2B - Form 3B: All transaction (income, expenditure and consumption) for the third and fourth weeks except what is collected in Forms 2A and 2B - Form 4: All transaction (expenditure and consumption) for last 6 months for Household expenditure on some selected item groups
- Form 5: Cash income and receipts received by household and type of tenure.

Cleaning operations

Data Editing, Coding and Capturing:
The first step of data processing activities was the training of 40 data editors/ coders and 20 supervisors by subject matter department staff members for the first round survey data. The data capturing (data entry) operation was carriedout using about 60 computers and as many data encoders. Similarly, the data processing activities of the second round survey were undertaken by about 60 editors/coders and 25 verifiers for about 85 days. Data entry operation took about 60 days using 125 computers and as many data encoders.

Data validation and cleaning activity was carried out by subject matter specialists and data processing programmers. The data cleaning and validity

Abstract

Poverty is a major concern and an important issue in Ethiopia. It is a chronic problem exacerbated by war, draught and inappropriate polices. According to recent but ad-hoc studies, the majority of the country's population id estimated to fall below the poverty line. Though, this is not supported by any statistical survey results, it is no doubt carried some truth considering the living conditions of the rural as well as the urban poor in this country.

As in the case of a number of African countries which undertook the Social Dimensions of Adjustment (SDA) program, the issue of welfare monitoring in Ethiopia arose as part of the Economic Reform Program (ERP) currently being undertaken in the country. The ERP specifically and strongly underlies the effects of the reform program on poverty and the analytical capacity of the government to monitor such effects. To this end, a Welfare Monitoring System (WMS) was set up by the government by mid 1994.

In view of the wider context of the problem, the establishment of the WMS, as justified above is envisaged to consist major elements such as establishing an information system that provides a continuous picture of the poverty scenario in the country, indicate the impact of reform programs on the level of household welfare, establish follow-up procedures on the various programs and activities targeted towards poverty alleviation and conduct regular statistical surveys to assess in particular the efficiency of the efficiency of targeted programs.

Welfare Monitoring Survey of 1995 was the first of its kind to be undertaken in Ethiopia covering both rural and urban area. The WMS places emphasis on six basic needs indicators. These are health, education, anthropometry, access to selected facilities/services, amenities and assets. However, the distribution of households by domestic expenditure was obtained form the Household Income, Consumption and Expenditure (HICES) in which the WMS data were classified by expenditure group. Moreover, the food expenditure and relevant per capita calorie intake were computed form the HICE survey and reported in the Household Income, Consumption and Expenditure Survey Bulletin.

Objectives of the Welfare Monitoring Survey - provides baseline data on existing poverty situation and establish a system of information gathering on relevant key indicators; - identify poor and vulnerable groups that could be the focus of targeted intervention programs; - undertake periodic surveys and researches to evaluate targeted programs; - assess the short and medium term effects of macroeconomics and sectoral policies and programs on the poor; - produce conclusive reports and suggestions needed for due attention by the government and concerned implementing agencies.

Geographic coverage

The 1995 (1987 E.C.) Welfare Monitoring Survey covered the population in the country on a sample basis excluding the non-sedentary population in Afar and Somali Regions. The survey population constitutes all the eleven administrative regions (i.e., Tigray, Afar, Amhara, Oromiya, Somali, Benishangul-Gumuz, Souther Nations and Nationalities Peoples (SNNP), Gambela, Harari, Addis Ababa and Dire Dawa) covering both rural and selected urban areas.

The survey was conducted on the basis of 12 rural and 15 urban households selected in each EA and a total of 11, 569 households (7,010 in rural and 4,559 in urban) all over the country were covered in the survey.

Analysis unit

• Households
• Individuals (including adult women aged 15 and above, children aged 5 and below)

Universe

The survey covered all population in the selected samples except residents of collective quarters, homeless persons and foreigners.

Kind of data

Sample survey data [ssd]

Sampling procedure

The 1995 Welfare Monitoring Survey covered both urban and rural areas of the country, except six zones in Somali Region and two zones in Afar Region . For the purpose of the survey, the country was divided into four categories. Urban areas were divided into two broad categories taking into account sizes of their population. Rural areas were also grouped into two categories.

Category I: Rural parts of eight regions were grouped in this category each of which was the survey domain (reporting level). These regions are Tigray, Afar, Somali, Benishangul-Gumz, Gambela, Harari, Addis Ababa and Dire Dawa.

Category II: In this category thirteen survey domains were defined by grouping contiguous rural parts of the zones or special weredas in Amhara, Oromiya, and SNNP Regions respectively. These were: a) Amhara I) North Gonder, South Gonder II) East Gojam, West Gojam and Agew Awi III) North Welo and Wag Himra, and IV) South Welo, Oromiya and North Shoa.

b) Oromiya I) East Welega, and Welega, II) Ilubabor and Jimma, III) North Shoa, West Shoa, IV) East Shoa, Arsi Bale and Borena, and V) East and West Hararge.

c) SNNP I) Keficho-Shekicho, Bench-Maji and yem, II) North Omo, Derashe and Konso, III) Grage, Hadia and Kembata-alaba-Timbaro, and IV) Sidama, Gedio, Amaro and burji.

Other than the 13 domains (reporting levels) defined in Category II, three additional domains have been constructed by combining basic domains from the two rural categories. These domains are: a) Rural Amhara b) Rural Oromiya and c) Rural SNNP

Category III: Ten large urban centers of the country were grouped in this group. Each of the ten urban centers in this category was the survey domain, for which separate survey results for major survey characteristics were reported.

Category IV: Urban centers in the country other than the ten urban centers in category III were grouped in this category and formed a single reporting level.

Other than the eleven domains (reporting levels) defined in Category III and Category IV, one additional domain, namely total urban (country level) can be constructed by combining the basic domains defined in the two categories.

All in all twenty four basic rural domains (reporting levels) including total rural (country level) were defined for the survey.

In addition to the above urban rural domains, survey results are to be reported at regional and country levels by aggregating the survey results for the corresponding urban and rural area.

Definition of the survey domains was based on both technical and resource considerations. More specifically, sample sizes for the domains were determined to enable provision of major indicators with reasonable precision subject to the resources that were available for the survey.

The sample selection scheme and sample size: a) Category I and Category II: A stratified two-stage sample design was used to select the sample in which the Primary Sampling Units (PSUs) were enumeration areas (EAs). Sample EAs from each domain were selected using systematic probability proportional to size; size being number of households obtained form 1994 population and housing census. A total of 620 EAs were selected from the rural part. Within each sample EA a fresh list of household was prepared at the beginning of the survey's filed work and for the administration of the survey questionnaire 12 households per sample EA were systematically selected. For the 1995 WMS urban centers in the country other than the ten urban centers were grouped in this category. There is one domain in this category, namely; Other Urban Centers. Other than the eleven domains (reporting levels) defined in category I and Category II, one additional domain namely, total urban (country level) can be constructed by combining the basic domains defined in the first two categories.

b) Category III: Stratified two-stage sample design was used to select the sample in which the PSUs were EAs. Sample EAs from each domain were selected using systematic probability proportional to size; size being number of household obtained form the 1994 population and housing census. In this category, a total of 220 EAs were selected. Within each sample EA, fresh list of households was prepared at the beginning of the survey's field work and for the administration of the survey questionnaire 15 households per sample EA were systematically selected.

c) Category IV: Three-stage stratified sample design was adopted to select the sample from domains in category IV. The PSUs were urban centers selected using systematic probability proportional to size; size being number of households obtained form the 1994 population and housing census. The secondary sampling units (SSUs) were EAs which were selected using systematic probability proportion to size; size being number of households obtained form the 1994 population and housing census. Number of sample SSUs selected from each of the the sample urban centers was determined by proportional allocation to their household population from the census. Ultimately, 15 households within each of the sample EAs were selected systematically from a fresh list of households prepared at the beginning of the survey's field work the administration of the survey questionnaire.

Note: Distribution of sample units by domain (reporting levels) is given in Summary Tables A and B of 1995 Welfare Monitoring Survey report.

Mode of data collection

Face-to-face [f2f]

Research instrument

Using seven schedules, data was collected on basic population characteristics, health, education, housing, anthropometry, services and amenities available, distance from home to facilities/services, and profile of asset ownership.

Cleaning operations

Data Processing: Due to the magnitude and nature of the welfare monitoring survey special attention had been given to the data processing activities. Thus, a task force

Abstract

The level and distribution of poverty in Ethiopia is extensive. According to the results obtained from the 1995/96 and the 1999/2000 Household Income, Consumption and Expenditure Survey and Welfare Monitoring Survey of the Central Statistical Authority (CSA), about 44 percent of the total populations (45 percent in rural areas and 37 percent in urban areas) are found to be below poverty line. The country together with development partners, as a result, has put poverty reduction strategies high on the agenda and working firmly on the implementation program since the beginning of this decade. With firm dedication to reduce poverty, the government has prepared its poverty reduction program entitled “Sustainable Development and Poverty Reduction Program” in 2002. The strategy has been implemented over the past three years.

In light of the plan to reduce poverty over time, strong system of monitoring and evaluation has been put in place. Consequently, the issue of welfare monitoring in Ethiopia arose as part of the Economic Reform Program (ERP). The ERP specifically and strongly underlies to see the effects of the reform program on poverty and building the analytical capacity of the government to monitor and evaluate such effects. To this end, the government has established a Welfare Monitoring System in 1995 to oversight the following major activities: - establish an information system that provides a continuous picture of the poverty scenario in the country; - indicate the impact of reform programs on the level of household welfare; - establish follow-up procedures on the various programs and activities targeted towards poverty reduction; and - conduct regular statistical surveys to assess, in particular, the efficiency of targeted programs.

The CSA and WMU of Ministry of Finance and Economic Development (MoFED) has been the key actors in the M&E system in place by way of producing, analyzing and disseminating poverty related data and results. The program has been supported by the World Bank IDA Credit and the Norwegian Trust Fund.

In line with strengthening the established M&E system of the government, the CSA and WMU of MoFED had also submitted a financing requirement proposal (entitled: Strengthening Data Collection, Analysis and Dissemination on Poverty Monitoring and the MDGs) to the Development Assistant Group (DAG). Following the signing of the Memorandum of Understanding (MOU) between MoFED, CSA and the donors, various implementation activities have also been carried out.

In order to meet the data needs of the Welfare Monitoring System, the CSA has been conducting the two surveys that provide poverty related data: Household Income, Consumption and Expenditure Survey (HICES) and Welfare Monitoring Survey (WMS) since 1995/96. The HICE and WMS surveys provide crucially useful information for the designing and monitoring and evaluation of the country’s poverty reduction strategy: Sustainable Development and Poverty Reduction Program (SDPRP), the various socioeconomic policies and programs and hence monitor the progress towards meeting the Millennium Development Goals (MDGs). The two surveys are inseparable and provide basic data that reflect the standard of living of households, individuals and the society as a whole.

The WMS focuses on wider range of socioeconomic indicators that reflect the nonincome dimensions of poverty. The 2004 survey have been improved to accommodate users needs as much as possible. New features in the current survey include shocks and coping mechanisms, HIV/AIDS related information, estimates of orphan and foster children, major prevailing diseases, plot size and other additional variables.

The main objectives of the WMS are to provide data that enable understand the nonincome aspects of poverty and has the following objectives: - to assess the level, extent and distribution of non-income poverty; - helps assess the quality of life of households/individuals; - to provide basic data that enables design, monitor and evaluate the impact of socioeconomic policies and programs on households/individuals living standard; - provide basic indicators on households’ and individuals’ living standard with respect to basic needs including: -> Education, Health, Child nutrition and care, Access to and utilization of basic facilities, Housing and housing amenities (drinking water, sanitation, energy, etc.), Household assets, Selected indicators of living standard, Vulnerability (Shocks and coping mechanisms, Food security, etc.), and HIV/AIDS and basic population characteristics.

Geographic coverage

The year 2004 Welfare Monitoring Survey covered all rural and urban area of the country except the non-sedentary areas in Afar and Somali Regional States. Excluded are three zones of Afar Region, six zones of Somali Region and all zones of Gambella Region.

Analysis unit

• Households
• Individuals (including adult women aged 15 and above, children aged 5 and below)

Universe

The survey covered households in the selected samples except residents of collective quarters, homeless persons and foreigners.

Kind of data

Sample survey data [ssd]

Sampling procedure

Sampling Frame The list of all households obtained from the 2001/02 Ethiopian Agricultural Sample Enumeration (EASE) is used as a frame to select the sample EAs in the rural areas of the country. In urban areas, on the other hand, the frame, consisting of households by EA obtained from the 2004 Ethiopian Urban Economic Establishment Census (EUEEC), is used to select sample sites for the 2004 welfare monitoring survey. The frame from which sample households were selected was based on a fresh list of households taken at the beginning of the survey period in each of the selected urban and rural EAs.

Sample Design For the purpose of the survey the country was divided into three broad categories including rural category, major urban centers category and other urban centers category.

Category I - Rural: This category consists of the rural areas of 58 zones and special Weredas, which are considered as zones, in 7 regions of the country. Harari region and Addis Ababa City Counsil and Dire Dawa City council were treated independently. Each zone/special Wereda of a region was considered to be a survey domain (i.e. reporting level) for which the major findings of the survey are reported. Rural Harari, Addis Ababa and Dire Dawa each were considered as separate reporting levels.

Category I totally comprises 61 reporting levels. A stratified two-stage cluster sample design in which the primary sampling units (PSUs) were EAs was used to select samples. Twelve households per sample EA were selected as a second Stage Sampling Unit (SSU) to which survey questionnaire finally were administered to the members of sample households.

Category II - Major urban centers: In this category all regional capitals (except Gambella region) and four other major urban centers that have relatively larger population sizes were included. Each of the 14 urban center in this category is taken us a reporting level. Since there is a high variation in the standards of living of the residents of these urban centers (that may have a significant impact on the final results of the survey), each urban center was further stratified into the following three sub-strata.

Sub-stratum 1: Households having a relatively high standard of living Sub-stratum 2: Households having a relatively medium standard of living Sub-stratum 3: Households having a relatively low standard of living

In this category too, a stratified two-stage cluster sample design was adopted to select the primary sampling units (the EAs) and the Second Stage Unit. Allocation of sample EAs to a reporting level among the above mentioned strata were accomplished in proportion to the number of EAs in each stratum. Sixteen households from each of the primary sampling units (EAs) in each reporting level were then selected as a Second Stage Unit (SSU).

Category III - Other urban centers: Urban centers in the country other than those under category II were grouped under this category. A domain of other urban centers is formed for each region (excluding Gambella region) and seven reporting levels were formed in this category. However, there is no domain in category III for Harari, Addis Ababa and Dire Dawa as they do not have urban centers other than those grouped under category II.

Unlike the above two categories, a stratified three stage cluster sample design was adopted to select samples from this category. The primary sampling units were urban centers and the second stage sampling units were EAs. Sixteen households from each of the selected EAs were finally selected as a third stage sampling unit.

Sample Size and Selection Scheme Category I: A totally of 2,016 EAs and 24,192 households were selected from this category. Sample EAs of each reporting level were selected using Probability Proportional to Size (PPS) with systematic sampling techniques; size being number of household obtained from the 2001/2 Ethiopian Agricultural Sample Enumeration. Twelve households per EA were systematically selected from the fresh list of households prepared at the beginning of the survey.

Category II: In this category 485 EAs and 7,760 households were selected. Sample EAs from each reporting level in this category were also selected using probability proportional to size systematic sampling; size being number of households obtained from the 2004 EUEEC. Sixteen households in each of the selected EAs were systematically selected from the fresh list of households prepared at the beginning of the survey.

Category III: One hundred twenty-seven urban centers, 275 EAs and 4,400 households were selected in this category. Urban