The National Statistical Office (NSO) first conducted the Household Expenditure Survey in 1957. Comparable survey known as "The Socio-Economic Survey was carried out in 1968-1969 and was repeated every five years. In 1986, due to the rapid economic expansion and the importance of the survey in order to set the anti-poverty policy, the Ministerial Cabinet had passed an approval on September 8, 1987 for the NSO to carry out the survey for every two years. The survey actually took place in 1988 and the 1990 survey is the tenth survey of this kind.
The survey provides information relevant in setting up the anti-poverty policy of the country. Its primary objective is to collect information on household income and consumption expenditures, changes in assets and liabilities, housing characteristics as well as the ownership of some durable goods.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Version 2 (May 2011). In this version, the datasets r3, r4, r6 to r10 were restructured from wide to long formats.
The scope of the 1990 SES includes:
HOUSEHOLD: household composition, income of household members, housing facilities, saving and debts, and expenditures for goods and services
The survey covered all private, non-institutional households residing permanently in municipal areas, sanitary districts and villages. Excluded from the survey are population living in transient hotels and rooming houses, boarding schools, military barracks, wats, hospitals, prisons and other such institutions, as well as households of foreign diplomats and other temporary residents.
Producers and sponsors
National Statistical Office
A stratified two-stage sampling was adopted for the survey. Greater Bangkok Metroplitan Area and region were constituted as strata. The primary and secondary sampling units were blocks for municipal areas and villages for non-municipal areas and private households, respectively.
Groups of provinces in each region and Greater Bangkok Metropolitan Area (Bangkok Metropolis, Nonthaburi province, Pathum Thani province and Samut Prakan province) constitute strata. There were altogether 5 strata i.e., Greater Bangkok Metropolitan Area, Central (excluding Greater Bangkok Metroplitan Area), North, North-East and South.
Selction of the Primary Sampling Unit
Each stratum was divided into three parts according to the type of local administration, namely, municipal areas, sanitary districts and non-municipal areas outside sanitary districts. The sample selection of blocks and villages were performed separately and independently in each part by using probability proportional to the total number of households.
Selection of Secondary Sampling Unit
Private households are our ultimate sampling units. A new listing households was made for every sample block and village to serve as the sampling frame, A systematic sample of 15 private households were selected from each of sample blocks, while 9 and 7 private households were selected from each of sample villages in sanitary districts and non-municipal areas outside sanitary districts, respectively.
The total number of sample private households selected for the enumeration is 15,920.
Sample weights for the household data were computed as the inverse of the probability of selection of the household. The weight variable for households is a32, found in the data file r1 - Household characteristics, household head.
Dates of Data Collection
Data Collection Mode
The collection of data was conducted under the direction of the NSO Field Operation Division. The interviewers, who were permanent members of the Field Operation Division working out of the NSO provincial branch offices, were sent out to interview the heads of the sample households or other household members.
Supervisors were expected to assist interviewers if problems occurred.
Data Collection Notes
All sample households were divided into 12 equally representative sub-samples, and each sub-sample household group was interviewed within one month. All households residing in sample blocks and villages were listed one or two weeks before the schedules interview periods in order to avoid too great a lapse of time between sample selection and interview.
Data Reference Period
During the first two weeks of the interview month, questions were asked about household composition, income of household members, housing facilities, saving and debts, and expenditures for goods and services as well. Various reference periods were used for collecting income and expenditure data. For all goods and services, data were obtained for the preceding month. In addition, for items usually purchased infrequently and also for sources of income, data were obtained for the past 12 months.
Limitations of the Data
It is generally thought that income and expenditure data are under-reported in household sample surveys, income to a greater degreee than expenditures, and the extent of under-reporting varies for income from different sources.For example, wage and salary earnings are probably much more accurately reported than profits from business. Households are likely to forget to report some expenditures made during the reference period of the survey and household members may be reluctant to report certain types of purchases, for example, alcoholic drinks, cigarettes and entertainment costs.
Non-money income was imputed by respondent. Own produced consumed by household and goods received free were valued at the local current retail prices. The rental value of owner-occupied home was the value that would have been received per month has the house been rented out to other households.
Permanent members of the Field Operatin Division working out of tne NSO prvincial branch offices
National Statistical Office
Two questionnaires were used:
- SES 2: for household composition, demographic and economic characteristics of household members; income and housing facilities
- SES 3: for household expenditures and 7-day food consumption
Stages at which data editing took place:
a) In the field:
- Each completed interview was subjected to a thorough field edit, followed by a follow-up interview if the information was found to be incomplete or internally inconsistent.
- When the household balance sheet which compared total money "disbursements" with total money "receipts" for the preceding month was more than 15 per cent out of balance, a revisit to the household was expected to reconcile the difference.
- Members of the Central Office Staff conducted periodic visits to the field to review questionable reports and clarify data collection procedures.
b) At the Central Office by subject matter staff of the Economic Statistics Division
c) At the Central Office by the Data processing Operations Division
d) During data entry
e) Prior to tabulations
Descriptive information was coded numerically for computer processing. All annual expenditure and income values were converted to a one-month base by dividing annual values by 12. For 7-day food consumption, values were multiplied by 4.3 which was the average number of weeks per month (52 weeks/12 months = 4.3 weeks/month). Income from farm or non-farm enterprises was calculated on the basis of total annual value of production less operating expenses. From this estimate, the value of products held or withdrawn for household consumption was subtracted to arrive at an estimate of money income.
Details on imputation and derivation for each expenditure group can be found in Socio-Economic Survey 1990 - Data Dictionary.pdf.
Data processing involves two processes:
1. Editing and coding. Reviewed questionnaires were again reviewed by the Data Processing Operations Division for completeness and consistency in details. Descriptive information was coded numerically for computer processing. Conversions and estimations were also done at this stage.
2. Tabulation: All information to be tabulated was keyed to tape for computer processing. Before tabulations were prepared, all raw data were edited by computer for final review and correction to ensure that data were all clean for further tabulation. Tables reflect averages and percentage distributions.
Estimates of Sampling Error
Formulas for coefficient of variations (CV) for totals and averages for characteristics (Y) of households is presented in Chapter 2 of the Report of the 1990 Household Socio-Economic Survey.
Section 16 of The Statistics ACT B.E. 2550 (2007), 30 August, B.E. 2550 guarantees that "anyone performing duties in an agency or the National Satistics Office shall not use the personal information provided by the owner or given in the questionnaire for any purpose other than the statistical analysis or research."
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including acronym and year of implementation)
- the survey reference number
- the source and date of download
Thailand National Statistical Office (TNSO). Thailand Household Socio-Economic Survey 1990. Dataset downloaded from http://web.nso.go.th on [date].
National Statistical Office
Disclaimer and copyrights
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 Document ID
Conversion to DDI
Date of Metadata Production
DDI Document version
Version 2 (May 2011). In this version, the datasets r3, r8 to r12 were restructured from wide to long formats.