Household expenditure survey
- Registration entry for subjects
- Contents
- Time
- Reliability and security
- Comparison
- Access to information
0. Registration entry for subjects
0.1 Name
Household expenditure survey0.2 Subject area
Consumption0.3 Responsible authority; office, division, person etc.
Finnbogi GunnarssonIndex Department
Statistics Iceland
finnbogi.gunnarsson@hagstofa.is
telephone 528 1209
0.4 Purpose and history
A continuous expenditure survey commenced in the beginning of the year 2000, and has been carried on since then. Previous surveys were carried out in 1939/1940, 1953/1954, 1964/1965, 1978/79, 1985/86, 1990 and 1995.Comparable surveys are carried out in other countries.
The main purpose of the household expenditure survey is to gather information on the consumption expenditures of households for the expenditure base of the consumer price index. It provides important insight into changes to patterns of consumer spending and information on household expenditure and its composition according to various social and economic factors.
0.5 Users and application
In addition to being the base of the consumer price index, its value is extensive:- in evaluating private consumption for national accounts
- in surveys, in the social and economic fields
- in market research
- in international comparisons of consumption.
0.6 Sources
The expenditures survey is a sample surveySources can be divided into three categories: Interviews with participants, household expenditure diaries, and information from the National Register of Persons, tax records and the Land Registry.
0.7 Legal basis for official statistics
Current act no. 12/1995 regarding the consumer price index stipulates that Statistics Iceland should carry out a household expenditure survey at intervals of no more than five years.0.8 Response burden
Those in the sample can choose not to respond.Respondents keep accurate household expenditure diaries for two weeks. They can choose to hand in receipts from stores, which eases the response burden.
Moreover, respondents are interviewed at the onset of the survey to gather information about the household. At the end of the diary recording period questions are asked on major infrequent expenditure items over the previous three months. The response burden for participants is therefore some.
In addition to this data information is gathered from the National Register of Persons, the tax register and the Land Registry, as well as information from telephone companies.
0.9 EEA and EU obligations
No formal treaties or rules. EEA regulation on the calculation of harmonized consumer price indices prescribes that expenditure surveys are carried out regularly.1. Contents
1.1 Description of content
The survey gives exact and itemized information on the consumption of households. The following can be found in the expenditure survey:- Average annual household expenditure divided into expenditure categories and disposable income.
- Housing. Residence in owned versus rented dwelling, type and size of dwelling.
- Possession of consumer goods (cars, household goods, telephones, etc.) and subscriptions to TV stations, newspapers etc.
- Size and type of household according to place of residence.
Average annual household expenditure divided into expenditure categories and disposable income are divided by:
- place of residence, Capital area, towns outside capital area (communities with more than 1,000 inhabitants), other communities (less than 1,000 inhabitants)
- type of household (singles, couples with and without children, single parents, others)
- quartiles of income
- quartiles of expenditure
Sample. Each year the sample consists of 1,200 households, chosen at random from family numbers of individuals in the National Register of Persons aged 18-74. Participants are all those living in the household of the chosen individual.
Data collection. In interviews with participants at the beginning of the survey information is gathered on, among other things, the number of household members, how they are connected, their identification numbers, education and employment. Participants keep household expenditure diaries for two weeks and note all expenses. Households are provided with special household expenditure diaries in which to note expenditures. If they are sufficiently itemized, receipts from stores can be used instead of noting each item in the booklet. After finishing the two weeks' diary keeping households in the capital area, Akureyri, Reykjanesbær and their surroundings are visited by interviewers while other households are sent a questionnaire. In these interviews / questionnaires participants are asked on, among other things, rare and/or more substantial expenses over a three month period, such as housing expenses, car related expenses, expenses due to travel abroad and purchasing of furniture and electrical goods. Questions are also asked on the size and type of house, as well as on household durable and home electronic good ownership and repayments of loans. In addition information from the National Register of Persons, tax records and the Land Registry is used.
1.2 Statistical concepts
The research unit is household. The sample is drawn on a random basis from the National Registry of Persons. Family identity number of people aged 18-74 are chosen irrespective of residence or marital status. Participants are all those living in the household selected.Consumption: Households' purchasing of good or service
Households: All individuals living under the same roof and running a common household while the survey was being carried out.
Equivalent scale: Differences in household size can make it difficult to interpret the results of expenditure surveys. One method for taking into account different household sizes involves the use of weightings to reflect their different compositions. Household expenditures need not increase in direct proportion to the number of members. Recalculating the household size by using equivalent scales attempts to take into account the fact that larger homes are more economical to run than small ones, and that expenditures on children are lower than on adults. Each individual in the household is given a specific weight according to age and household size, enabling expenditures to be compared among different types of households. International agencies such as OECD and Eurostat use equivalent scales. The OECD uses a scale which assigns the first adult individual in a household a weight of 1, other adults 0.7 and children 0.5. Eusrostats scale assigns the first adult in a household a value of 1, other adults 0.5 and children 0.3. Both these systems assume the same level of expenditure on young people who reach the age of 13 as on adults.
2. Time
2.1 Reference periods
Data is collected continuously and results are based on data collected three years in a row. The sample, which consists of 1200 households per year, is too small to allow sufficiently reliable and extensive results to be reached every year.Therefore results are based on three years' worth of data at the price levels of the last year. The first ones are based on data for the years 2000, 2001 and 2002 after which a three year moving average is used.
2.2 Process time
A continuous expenditure survey was commenced in the beginning of 2000 and in March 2002 the first results were used in the preparation of a new index base. Aggregate results for the first three years period were published in June 2004, the second in June 2005 and the third in January 2006. Ongoing is work on cutting down the process time of the survey. The intention is to publish every three years results in December the year after the periods end.2.3 Punctuality
Results are published according to the advance release plan at 9 am.2.4 Frequency of releases
Results are released annually, in December.3. Reliability and security
3.1 Accuracy and reliability
Results from sample surveys are always subject to some uncertainty, due to only part of the population being surveyed. The greater the level of itemization used, the greater the sample required to make the results reliable. Results on expenses also vary in reliability according to the type of good or service. It can be assumed that information on frequent expenses is more reliable than information on rare ones. In many expense categories most households have no expenses, while a few have significant expenses. Examples of this are car purchase and purchase of furniture and household goods.3.2 Sources of errors
Errors in data can occur in a number of ways, for example:- Sampling errors (coverage errors, non-response errors).
- Measurement errors (due to interviewers, inaccurate forms, inaccurate answers or differing methods in data collection)
- Processing errors (errors when entering data, errors in the classification of data, etc.)
- Errors in data or missing information so values must be estimated.
- Errors in official registers.
Coverage. The population of the expenditure survey is private households in Iceland where someone in the household is younger than 75 years. In the National Register of Persons it is not possible to distinguish households but nuclear families are specified. Nuclear families as well as singles have a family number. The sample frame of the survey is family numbers of individuals in the National Register of Persons aged 18-74. The sampling probability is not the same for all households. Households with a number of individuals aged 18-74 have a greater probability of being sampled. In the processing of data this is taken into account.
Non-response. A high level of non-response can cause errors in the results of the survey if the non-response group deviates from those who participate in the survey. Experience from previous surveys has shown that there is large difference in response according to the type of household in question. Those who are least likely to participate are singles, while it has been easiest to get couples with children to participate. In the processing of the survey results are adjusted according to non-response.
Processing. The main problems with processing are that results must be corrected for different sampling probabilities as well as non-response. Moreover, answers are often missing to individual questions in which case values must be estimated for gaps. Median values of the expense categories of those who do respond are usually used. Expenses are also estimated with regard to the household's other answers if possible.
3.3 Measures on confidence limits/accuracy
The higher the level of itemization in the response, e.g. according to residence, type of household, etc., the larger the sample needed to make conclusions reliable. In results from the expenditure survey sums and percentages are marked with an asterisk (*) if the relative standard error is greater than 20%. 4. Comparison
4.1 Comparison between periods
The implementation of the surveys has changed somewhat from when they began. The diary recording period been reduced from one year in the first two surveys to two weeks, while data on major infrequent expenditure items has been especially collected. Sampling methods have also changed between surveys. Due to changes to the classification framework there are some difficulties in comparing the results of the newest surveys to those of older surveys.4.2 Comparison with other statistics
Results of the expenditure survey are compared to other sources where possible, such as import statistics, industrial production statistics, agricultural production statistics and industrial turnover figures. The results are also compared with national accounts.4.3 Coherence between preliminary and final statistics
Preliminary statistics are not published.5. Access to information
5.1 Forms of dissemination
- News, released on Statistics Icelands website
- Statistics, categorised statistical web tables
- Statistical Series, Hagtíðindi. Results are published annually in a special issue of the Statistical Series, where there is extensive information on the implementation, processing and reliability of the survey. Results are published in tables.
5.2 Basic data; storage and usability
The source material is stored in digital form by Statistics Iceland. No access is allowed to the data itself but it is possible to have it processed specially.5.3 Reports
In the report Household Budget Survey 1995 (Statistics of Iceland III, 50) the results of the expenditure survey of 1995 are explained.5.4 Other information
Further information is provided by:Finnbogi Gunnarsson
telephone 528 1209
© Hagstofa �slands, �ann 6-12-2012