SWZ_1995_HIES_v01_M
Household Income and Expenditure Survey 1995
Name | Country code |
---|---|
Kingdom of Eswatini | SWZ |
Income/Expenditure/Household Survey [hh/ies]
The first Household Income and Expenditure survey was conducted in 1985 with the assistance of the German Government, the Economic Commission for Africa and the Commonwealth Fund for Technical Co-operation. The 1995 survey was carried out from November 1994 to October 1994, and it covered a total of 6392 households. The main objectives of the HIES 95 were in the first place to provide basic data required for policy making at national and regional levels as well as for different sectors.
The Swaziland Household Income and Expenditure Survey 1995 (the SHIES 1995) is the second survey of this nature to be carried out by the Central Statistical Office (CSO). The main objectives of the SHIES 95 were in the first place to provide basic data required for policy making at national and regional levels as well as for different sectors. Secondly the survey would facilitate the determination of needs or in the establishing of targets which can be identified as obtaining:
Sample survey data [ssd]
The Household Income and Expenditure Survey 1995 covered the following topics:
National
Name |
---|
Central Statistical Office |
The sample design is based on the nation-wide Master Sample, covering both urban and rural areas (see references 2 and 3). The institutional population is not included in the survey. These people do not live in private households but in institutions like hospitals, hostels, military or police barracks, prisons etc.
A stratified two-stage sampling design is used with the Census 1986 enumeration areas (EAs) as the primary sampling unit and homesteads as secondary sampling unit. The number of sampled EAs were 216 of the total 1079 EAs. They were selected by using PPS (probability proportional to size, i.e. number of homesteads). All together 6350 homesteads were sampled with systematic random sampling within the EA. Urban areas are slightly over sampled.
As the distribution of households over the time is rather uneven, there are reasons to believe that it can cause bias. The participating households in December are much higher, with a risk of overestimating the annual expenditures as much more money is spent during Christmas time. Anyhow, most of the households participating the last two weeks of November 1994 through the first two weeks of December have been incorrectly coded as participating in December 1994. This weighting procedure will, however, not take care of the bias caused of non-response where the reason for non-response is correlated with particular expenditures. In January few households are participating due to vacations. Expenditures for vacations will be underestimated even if we re-weight for the period. Those actually measured in January were probably not on vacation and therefor had no expenditures for that.
Start | End |
---|---|
1995 | 1995 |
Fieldwork
The fieldwork was carried out by 4 supervisors and 90 interviewers. Before data collection started the 4 supervisors and 30 interviewers were trained f0r 10 days, 3 of these days were used for a mini pilot survey, when selected households were interviewed and kept records for 2 to 3 days. After 2 months a further 60 interviewers were recruited. No formal training on the SHIES were given. They were supposed to act as assistant interviewers collecting data for the Child Mortality Study in FORM A.
The supervisors came into Office once a week. And most of the interviewers once a month.
Method of data collection
The selected households participated for a period of four weeks. During the initial interview the interviewer collected socio-economic details of every member of the household. Some housing characteristics were recorded at the same time. The household was then instructed on how to keep a record of daily expenditure or other consumption, and informed that regular visits would be paid thereafter.
The main instruction was to make the household to fill in FORM B themselves. However, many households could not fulfil this ask due to illiteracy. In these cases some person in the household, a schooll child, or a neighbor kept records in a notebook and the interviewer transferred these records to FORM B every second or third day.
At the last day of the recording period the interviewer asked for details of income received during the previous month. This applied only to household members aged 12 years or more.
Coding and Editing
The editors coded occupation, industry and consumption transactions. The rest of the questions in the questionnaires were pre-coded.
Data Entry
The software used for data entry was the IMPS Version 3.
The IMPS data entry program and validation program caused a lot of problems. It created duplicates households and individuals and strangely enough these duplicates seemed to be created at random.
Other problem was a number of batches had disappeared after data entry. To solve this more problem, more than 30% of the FORM B had been re-punched/ re-entered. A lot of effort was used to clean the data, however, all problems could not be solved.
The use of the datasets must be acknowledged using a citation which would include:
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
DDI_SWZ_1995_HIES_v01_M_WB
Name | Affiliation | Role |
---|---|---|
Development Data Group | World Bank | Generation of DDI documentation |
2013-06-12
Version 1.0 (June 2013)