Abstract

The 2016/2017 Uganda National Household Survey (UNHS) is the sixth in the series of national household surveys conducted by UBOS. It is a follow-up of the 1999/2000, 2002/03, 2005/06, 2009/10 and 2012/13 UNHSs. The survey collected socio-economic data required for measurement of human development and for monitoring social goals, with particular focus on the measurement of poverty and unemployment for the Sustainable Development Goals (SDGs) and the National Development Plan II (NDP II) among other policy frameworks.

Objectives : The main objective of the survey was to collect high quality and timely data on demographic, social and economic characteristics of the household population; to inform as well as monitor international and national development frameworks.

The specific objectives of the survey were to: 1. Generate and provide socio-economic indicators for monitoring and tracking progress towards Uganda’s development goals; 2. Meet the data needs of key users including Ministries, Departments and Agencies (MDAs) of Government and Development Partners among others; and 3. Strengthen the technical capacity of UBOS to plan, conduct, process and analyze complex survey data in the long-term.

The data collected in the 2016/17 UNHS adds to the growing national database of socio-economic indicators.

Geographic coverage

• National coverage
• Urban and rural areas
• Sub-regions of Uganda:Kampala, Central I, Central II, Busoga, Bukedi, Elgon, Teso, Karamoja, Lango, Acholi, West-Nile, Bunyoro, Tooro, Ankole, and Kigezi.
• Districts: At the time of the survey there were only 112 districts, this number later increased to 122 districts.

Analysis unit

• Individuals
• Households
• Communities

Universe

The survey was based on the household population and excluded the population in institutions, refugee camps, forest reserves, police and army barracks, and other special areas.

Kind of data

Sample survey data [ssd]

Sampling procedure

The 2016/17 UNHS sample was designed to allow for generation of separate estimates at the national level, for urban and rural areas and for the 15 sub-regions of Uganda. At the time of the survey there were only 112 districts. This number later increased to 122 districts. A two-stage stratified sampling design was used. At the first stage, Enumeration Areas (EAs) were grouped by districts of similar socio-economic characteristics and by rural-urban location. The EAs were then drawn using Probability Proportional to Size (PPS). At the second stage, households which are the ultimate sampling units were drawn using Systematic Random Sampling.

A total of 1,750 EAs were selected from the 2014 National Population and Housing Census (NPHC) list of EAs which constituted the Sampling Frame. The EAs were then grouped into 15 sub-regions, taking into consideration the standard errors required for estimation of poverty indicators at sub-regions and the rural-urban domains.

Mode of data collection

Computer Assisted Personal Interview [capi]

Response rate

Of the total 17,320 households selected for the 2016-2017 UNHS sample, 15,672 households were successfully interviewed, giving a response rate of 91 percent. The response rate was higher in rural areas (93%) compared to urban areas (88%).

Sampling error estimates

The estimates from a sample survey are affected by two types of errors: non-sampling errors and sampling errors. Non-sampling errors usually result from mistakes made during data collection and capture and those include misunderstanding of the questions, either by the respondent or by the interviewer and by capture of wrong entries. Such errors were controlled through rigorous training of the data collectors and through field spot-checks undertaken by the supervisors at the different levels.

On the other hand, sampling errors (SE) are evaluated statistically. The 2016/17 UNHS sample is one of the many possible samples that could have been selected from the same population using the same sampling design. Sampling errors are a measure of the variability between all possible samples that would yield different results from the selected sample. Sampling errors are usually measured in terms of the standard error for a particular statistic such as the mean, percentages, etc. The Tables in Appendix III present standard errors and Coefficients of Variations (CVs) for selected indicators at national, rural urban and sub-regional levels.

Abstract

The 2016/17 Uganda National Household Survey (UNHS) is the sixth in a series of national household surveys that Uganda Bureau of Statistics (UBOS) has undertaken. The survey collected information on socio-economic characteristics at both household and community levels. The main objective of the survey was to collect high quality data on demographic and socio-economic characteristics of households for monitoring Uganda’s development performance of key indicators in the various sectors. The 2016/17 UNHS comprises four (4) modules. Those are the Socio-Economic, Labour Force, Community, and Market price modules. The main findings are based on the four modules and include trends of several indicators on Education, Health, Household Expenditure and Poverty, Food security, Income and loans, Information and Communication Technology, Vulnerable Groups, Community Characteristics and Non-crop household enterprises, presented at national, rural-urban, regional and sub-regional levels. The survey collected much more information besides what has been included in the main findings. Therefore, UBOS calls upon all stakeholders to utilize the wealth of data collected and availed over the years to undertake in-depth empirical analysis so as to better inform future policy debate.

Geographic coverage

National coverage

Analysis unit

The UNHS 2016/17 had the following units of analysis: individuals, housheholds, and communities.

Universe

The survey covered all de jure household members (usual residents), all currently employed and unemployed persons aged 5 years and above, resident in the household.

Kind of data

Sample survey data [ssd]

Sampling procedure

The 2016/17 UNHS sample was designed to allow for generation of separate estimates at the national level, for urban and rural areas and for the 15 sub-regions of Uganda. At the time of the survey there were only 112 districts. This number later increased to 122 districts. A two-stage stratified sampling design was used. At the first stage, Enumeration Areas (EAs) were grouped by districts of similar socio-economic characteristics and by rural-urban location. The EAs were then drawn using Probability Proportional to Size (PPS). At the second stage, households which are the ultimate sampling units were drawn using Systematic Random Sampling. A total of 1,750 EAs were selected from the 2014 National Population and Housing Census (NPHC) list of EAs which constituted the Sampling Frame. The EAs were then grouped into 15 sub-regions, taking into consideration the standard errors required for estimation of poverty indicators at sub-regions and the rural-urban domains. In addition to the sub-regions, the other sub-groups that were considered during the analysis of the 2016/17 UNHS include the Peace and Recovery Development Plan (PRDP) districts and Hard-to-reach areas such as the mountainous areas. The survey targeted to interview 10 households per EA, implying a total sample of 17,540 households. Prior to the main survey data collection, all the sampled EAs were updated by listing all the households within their boundaries.

Mode of data collection

Computer Assisted Personal Interview [capi]

Research instrument

The UNHS 2016/17 adminstered four questionnaires including: Socio-Economic, Labour Force, Market Prices, and Community. All questionnaires and modules are provided as external resources in this documentation.

Response rate

Out of the total 17,320 households selected for the 2016/17 UNHS sample, 15,672 households were successfully interviewed, giving a response rate of 91 percent. The response rate was higher in rural areas (93%) compared to urban areas (88%).

Sampling error estimates

The estimates from a sample survey are affected by two types of errors: non-sampling errors and sampling errors. Non-sampling errors usually result from mistakes made during data collection and capture and those include misunderstanding of the questions, either by the respondent or by the interviewer and by capture of wrong entries. Such errors were controlled through rigorous training of the data collectors and through field spot-checks undertaken by the supervisors at the different levels. On the other hand, sampling errors (SE) are evaluated statistically. The 2016/17 UNHS sample is one of the many possible samples that could have been selected from the same population using the same sampling design. Sampling errors are a measure of the variability between all possible samples that would yield different results from the selected sample. Sampling errors are usually measured in terms of the standard error for a particular statistic such as the mean, percentages, etc. The Tables in Appendix III present standard errors and Coefficients of Variations (CVs) for selected indicators at national, rural-urban and sub-regional levels.