The 2000-2001 Household Income and Expenditure Survey is the fifth survey of its kind conducted in Tonga. The last one was carried out in 1992/93, and the results were used in 1995 to rebase the Consumer Price Index (CPI).
The Household Income and Expenditure Survey 2000/01 (HIES) was carried out by the Tonga Statistics Department over a 12-month period, from October 2000 to September 2001. The survey collected detailed information on income and expenditure from more than 1,600 households. All adult members age 15 years and over of these households kept a detailed diary of daily expenditure over a two-week period.
The survey had three main objectives:
• To provide updated information for the expenditure item weights for the CPI;
• To provide some data for the components of National Accounts; and
• To provide information on the nature and distribution of household income and expenditure for planners, policy makers, and the general public.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
The 2000-2001 Household Income and Expenditure Survey covered the following topics:
- Household Identification Particulars
- Household Composition
- Value of Output and Payments for Goods and Services for the Last 12 Months
- Sources of Household Income and Receipts for the Last 12 Months
- Household Consumption and Expenditure
- Daily Expenditure on Food (2 weeks)
Producers and sponsors
Kingdom of Tonga
Secretariat of the Pacific Community
In designing the sample for the survey, an initial simple allocation was made of 10 percent of all households in Tonga. This resulted in an initial sample of 1640 households, made up 1093 households for Tongatapu, 276 for Vava'u, 149 for Ha'apai, 83 for 'Eua, and 39 for the Niuas. To allow for seasonal effects, the survey was carried out over four rounds, covering a 12-month period. This meant that the sample size for each round was 410 households.
There were various stages to the selection process. Within each island division, the required sample was allocated to districts in proportion to their sizes. A sample of census blocks was then selected systematically within each district, using a sampling interval of 1 in 9. Within selected census blocks, the required size of sample was selected systematically from the list of all households.
Due to financial limitations, some of the small islands were not included in the survey at all, or only included in certain rounds. For example, the two main islands of the Niuas were only to be included in one round each. In the Ha'apai group, two of the islands (Foa and Ha'afeva) were only to be covered in two rounds, while three others (Ha'ano, 'Uiha and Nomuka) were only to be considered for one round.
Dates of Data Collection
Data Collection Mode
Data Collection Notes
The fieldwork was carried out by about 25 specially selected interviewers, supported by a maximum of 10 senior staff of the Statistics Department who acted as supervisors (including the two out-based staff who work in Vava’u). The interviewers were recent school leavers who had completed Form 6 and who held a Pacific Secondary School Certificate (PSSC). There was a two-week training course for supervisors, followed by a two-week training course for interviewers (also attended by the supervisors). All field staff had to swear and sign an oath of confidentiality.
In order to take adequate account of seasonal effects, the survey was spread over four rounds throughout the 12-month period. The survey was publicized by means of TV and radio. Fieldwork for each round lasted about a month. This involved an initial week for listing households in the selected census block, selecting the required number of households for interview, and making an initial contact with the selected households to drop off the diaries. The households then had two weeks for completion of the diaries, and during this period the interviewers would call in from time to time to check that the recording of daily expenditures was proceeding satisfactorily.
They would also use the occasion of their visits to collect some of the information required for the income questionnaire. They would then make a final visit to the households to collect the diaries. After checking the questionnaires, they passed them to their supervisors.
The interviewers were employed on a part-time basis, and were released after each round of fieldwork. Inevitably this created some problems, in that some of them managed to find full-time jobs and could not return to work on subsequent rounds of the survey. An additional five school leavers had been trained initially to form a reserve pool for the survey, but this supply of reserves had been exhausted by the end of the second round. It was therefore necessary to recruit and train some more school leavers to fill in for those who left at later stages of the survey.
When designing the survey, a decision had been made to only visit the Niuas in the first and fourth rounds, because of the very high travel costs involved. Fieldwork went according to plan in all of the island divisions with the exception of Ha’apai. The selected sample had initially been spread quite well around the Ha’apai group, but poor weather and lack of direct sea transport from Tongatapu meant that it was difficult to get to the southern islands in the group. A decision was therefore made to increase the sample on the more accessible northern islands in the group.
Two questionnaires (HIES-1 and HIES-2) were used to collect the data required for this survey:
HIES-1: Household income survey. This contained a cover sheet and three sections:
- Section A: Household composition and characteristics. This section collected information on the general demographic characteristics of members of the household selected for the survey, including relationship to head of household, sex, age, marital status, number of days worked last week, main activity, employment status, educational attainment, occupation and industry.
- Section B: Value of output and payments for goods and services for the last 12 months. This section collected information on goods produced for sale in the local market or for export, and goods produced for home consumption. Information was collected separately for the following items: agricultural products; livestock; poultry and poultry products; fish and fish products; handicrafts; and rent for property.
- Section C: Sources of household income and receipts for the last 12 months This questionnaire collected information on the source and amount of income received by income earners, separately for cash and non-cash income.
HIES-2: Household expenditure survey. This booklet contained instructions for completing the diary, a listing of items that people might have bought, and four sections:
- Section A: Daily expenditure during two weeks starting Saturday. There was a separate page for each day, and respondents were asked to give a description of each item obtained and the quantity. Alongside were three separate columns for recording the value of items that were paid for in cash, bought on credit, or where no cash was paid.
- Section B: Repairs on house during these two weeks. The value of all house repairs carried out over the two-week period was to be recorded here, unless it had already been recorded elsewhere in the diary. Again, there were three columns for cash, credit, and no cash paid.
- Section C: Rent for house If not already recorded elsewhere, this section was for recording the rent paid for the house, and the frequency of payment. In cases where a person lived in their own house, the person was asked to estimate the likely rental value if they had to rent the property.
- Section D: Other big payments during the last 12 months Any large payments (T$50 or more) made during the previous 12 months were to be recorded here. Examples were given of the types of payment to be recorded, and the layout was similar to that in Section A above.
All data entry, including editing, edit checks and queries, was done using the Access software package, with some technical assistance from the Secretariat of the Pacific Community (SPC). Preparation of the final tables was done using the Pivot Table facility in Excel.
When the data were being processed, it became clear that certain sections of the questionnaire gave rise to problems. Some of these problems are documented in Appendix C (refere to the final report). One major problem related to the treatment of the third column of expenditure figures in the two-week diary, under the heading ‘No cash paid’. As explained in Appendix C (refere to the final report), this third column of the two-week diary seems to contain a mix of different items: home produced items that have been consumed by the household;gifts received; and items bartered. It is impossible to disentangle these ‘no cash paid’ items, because no code was used to distinguish these different types of expenditure. This gives rise to two problems. Quite apart from the different basis used for valuing home produce (the income data is based on net income values after subtracting labour, transport and other costs from farm-gate values, while the expenditure data is based on market prices) it is not possible to directly compare the figures for consumption of home produce from the two sources because of this entanglement with gifts and barter in the diary. Similarly, the gift element in the diary cannot be compared with the data on ‘gifts, assistance, relief received in kind’, which was collected in Section C of the income questionnaire.
Initially a scheme was devised by SPC to attempt to get round this ‘gift’ problem. In the case of income, the larger of the two values ‘gifts received’ and ‘diary-no cash paid’ was taken as being the gift income. In the case of expenditure, the difference between these amounts was taken if ‘gifts received’ was greater than ‘diary-no cash paid’, while a value of zero was taken for gifts if ‘gifts received’ was not greater than ‘diary-no cash paid’. The SPC scheme also allowed for the inclusion in the expenditure aggregate of the net value of the home production consumed by the household, taken from the income questionnaire, although this category would seem to be covered already by the inclusion of the ‘no cash paid’ column from the diary.
Estimates of Sampling Error
Sampling errors refer to those errors that are implicit in any sample survey, where only a portion of the population is covered. Non-sampling errors refer to all other types of error. These can arise at any stage of the survey process. Examples of activities that are likely to increase the level of non-sampling error are: failing to select a proper sample, poor questionnaire design, weak field supervision, inaccurate data entry, insufficient data editing, or failure to analyze or report on the data correctly. If a census of all the households in Tonga were carried out, there would be no sampling error (but probably increased non-sampling error).
Note: Detailed sampling error descriptions and standard error tables are available in the 2000-2001 HIES final report.
Tonga Statistics Department
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 country, acronym and year of implementation)
- the survey reference number
- the source and date of download
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.