HIES data is an analytical tool that many countries use to gather information on household consumption on a regular basis to monitor changes in the level of poverty and analyse the well-being of the people and the economy. The main purpose of this survey was to collect information on the economic well-being of the population. Specifically, the results will be used to bring about improvements in the national accounts, consumer price index and the vulnerability and poverty statistics of the country.
The report on the Household Income and Expenditure Survey (HIES) of the Maldives 2002-2003 presents an analysis of findings from the survey. This is the first nationwide HIES conducted in the Maldives. The survey was conducted during the period September 2002 to June 2003 in four quarterly rounds during the months of September and December in 2002 and March and June in 2003. The survey was designed as a scientific random sample with separate strata for Male' and the five development regions. It covered the capital Male' and forty islands from the Atolls and includes data for 834 households.
As HIES data provide a snapshot of the socio-economic situation of the households prior to the Tsunami, it would also serve as significant baseline data in the future assessments and comparisons in this regard.
The tables presented in the report provide the data by standard classifications that include income quintiles, income classes, and daily per capita expenditure classes, cross-tabulated by demographic variables such as the size of household, number of earners, proportion of expenditure on different expenditure categories. To the extent possible, all information is presented separately for Male' and the Atolls and also at the regional level for the atolls and comparisons are made with the 1997/98 VPA.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
persons, households, expenditure items
HOUSEHOLDS: dwelling characteristics, housing expenditure, household operations.
INCOME: Wages and salaries, Wage income from non-resident earners, Business income, Imputed rent, Own production, Other Incomes, Property Income.
EXPENDITURE ITEMS: food, furnishing and furniture, clothing and footwear, housing, water, electricity, gas and fuels, medical and health expenses, transport and communcation, education, entertainment and sports, personal goods/personal care, miscellaneous goods and services.
Producers and sponsors
Department of National Planning
Republic of Maldives
United Nations Development Programme
United Nations ESCAP
Asian Development Bank
The country is divided into 5 development regions and 20 administrative atolls. Administrative atolls consist of 199 inhabited islands with clearly marked census enumeration blocks. The capital town of Male' has separate administrative status. HIES uses the area frame thus administrative and geographic structure of the country is taken as a basis to make the sample representative. Required data for sampling are obtained from the Population census 2000. Major characteristics of the frame are given below.
Atolls are too big to take as sampling unit, while size of islands in terms of the number of household varies even after some exclusion from merely 20 to 1500. Initially, it was thought to split some big islands and combine smaller to get evenly distributed area unit. Alternatively, census enumeration blocks are chosen to be primary sampling units from practical considerations. The size of enumeration blocks varies from 20 to 64 households.
Estimated mean of a characteristic from the sample survey results may deviate from the population mean resulting in a margin of sampling error. Relative error of the sample mean can be controlled by determining the sample size based on coefficient of variation.
Variance for income of households was estimated from the results of earlier surveys. Distribution of households by income group was skewed as expected due to the small number of households of high-income groups. After some extreme cases were excluded distribution came to be fairly normal with the coefficient of variation of 7.5%.
In order to produce the results with same precision at 95% level of confidence sample size n was estimated at 885 households (inversely, probability that sample mean calculated from results survey of 885 households will contain more than 7.5% of sampling error does not exceed 5%). In the course of the allocation process, actual number of households to be surveyed turned to be 880 and that much of deviation was accepted.
Stratification and allocation of sample
The purpose of the stratification is to divide the population into relatively homogeneous groups and thereby reduce the total variation by the margin of inter-group variation. Stratification allows proper allocation of sample in different groups and makes it more representative.
First, it was essential to treat Male' separately in the whole sampling procedure. The income opportunities and expenditure pattern in Male' is very much different from the rest of the country. It is also necessary to produce separate estimates for Male' like all other national surveys in Maldives. So, there are two domains of the survey namely: Male' and Atolls.
Stratification in Male' was done by wards and sub-wards to make the sample spread over different urban areas. Male has 5 wards out of which two wards each were split to 2 sub-wards due to their larger size.
Stratification in Atolls was done by development regions and by distance of islands to the atoll centre, which was aimed to group the islands of different regions remotely located from the atoll centres and those located nearby. It was believed that in most of the cases atoll centres are also economic centres providing more income opportunities and better access to different kind of social services. Islands remotely located from the atoll centres have limited facilities affecting on income and expenditure of households.
For grouping purpose, islands of each development region were listed in the ascending order of the distance from the respective atoll centres and a median value was located for the number of households. Then each atoll was divided to Central and Remote islands, where both groups were of more or less equal size in terms of the number of households. The average distance of Central islands in different region varied from 10 to 14 km and of Remote islands 27 to 37 km. Crossstratification by five development regions and two distance categories formed ten strata in the domain of atolls as shown in the chart.
Allocation of sample in domains was intended to make proportional to the number of households. However, it has not been possible due to the resource constraints. The survey cost per psu in Atolls was estimated from four to five thousand M.Rf., while survey cost in Male is limited to stationery because staffs do not get any extra allowances for working in Male.
Further allocation was made based on equal number of sample for all strata. The number of enumeration blocks as well as the number of households did not vary much, so the fixed number of samples over all strata resulted in sampling fraction ranging from 2% to 7% for enumeration blocks and 1.1% to 3.6% for households.
Allocation of sample in Male' strata is made on same principle. For each of the 7 strata equal number of 4 blocks are allocated with the rate of 10 households per stratum, which gives total number of 28 sample blocks and 280 households.
The sampling procedures are fully described in Appendix 1 of "Maldives Household Income and Expenditure Survey 2002-2003 - Final Report".
As mentioned above, the sample design for Male' included enumeration of the households during two rounds of the survey. It proved difficult to obtain the co-operation of some of the selected households for participation during the second round of enumeration and as a result, a larger than expected non-response was encountered. In total, 24 households, that is about 15 percent of the total, did not participate in their second period of enumeration. In the atolls, enumeration was not accomplished for only five households, or less than one percent of the total. In addition to the non-response of 29 households, in a small number of cases it was necessary to remove the household information during processing, basically because insufficient information was available. As a result, the final data set contains information for 834 households instead of the 880 in the design.
Weights can be incorporated in the programme and tables will be generated from weighted and unweighted data. Similarly estimation weight for atolls is calculated.
The weighting is fully described in Appendix 1 of "Maldives Household Income and Expenditure Survey 2002-2003 - Final Report".
Dates of Data Collection
Data Collection Mode
Data Collection Notes
A questionnaire will be administered to collect required data through fieldwork in four quarterly rounds starting from September 2002. The survey staffs will visit the sample area and complete the questionnaire prepared for this purpose. Before proceeding for questionnaire staffs will list all the households of enumeration block and draw a sample of prescribed number using SRSwor method.
A probability based sampling design has been prepared and enumeration blocks have been selected for all strata. Details of sampling design are described below. The survey questionnaire is annexed. The fieldwork is planned to integrate with the survey of small establishments, however sampling scheme for SES will be independent from HIES.
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
Department of National Planning. Maldives Household Income and Expenditure Survey 2002-2003, Ref. MDV_2002_HIES_v01_M. Dataset downloaded from [URL] on [date].
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.