EUROSTAT
Directorate E: Social and regional statistics and geographical information system
Unit E-2: Living conditions /
Doc. E2/HBS/128/01/EN
Working Group
HOUSEHOLD BUDGET SURVEYS
01-02 OCTOBER 2001
Eurostat-Luxembourg, Bech Building – Room Ampère
Point III/C of the agenda
Future Data Collection
1
Future data collection and development of HBS
1Background
The timeliness of the HBS has been (will be) improved significantly as we can present European results for HBS 1999 at the end of 2001. Compared to HBS1994 with final results in June 1999 this will definitely increase the interest for HBS. The main users of HBS-statistics are the National Accounts, the Price index statistics and policymakers within the area of Living conditions and Consumer protection. Within the Commission our main “clients” are Directorates General for Employment and Social Affairs and Health and Consumer Protection. There is also a considerable interest from the units in Eurostat working to present statistics on housing, housing expenditure, cultural expenditure and private expenditure on education.
The Household Budget Surveys are mainly national surveys with national users and the next Eurostat collection of data from the Household Budget Surveys depends to a large extent on the plans of the Member States. Both the frequency and the contents of the HBSs vary between the MS, although with the annual surveys in most countries and the common basic classification the output harmonisation procedure is fairly straightforward. Even if there is not a strong European political demand for developing particular topics in the HBS it may be interesting to share the country specific information and possibly to co-ordinate some efforts.
2Frequency of data collection
2.1Next data collection
At this stage all but five (EL, F, IRL, A and FIN) of the MS conduct or have plans to conduct annual surveys. Luxembourg, Germany and probably Portugal are changing into annual surveys (When?).
Table 1: Plans for Household Budget Surveys in the EU (October, 2001)
First survey: Taking place after 1945
X: Survey conducted during the calendar year
x: Survey conducted during 12 months spread over 2 calendar years
EU: The standardized Community Survey of 1963-1964 conducted in the original six Member States
A: Included in Eurostat Comparative Tables for reference year 1979 (10 countries)
B: Included in Eurostat Comparative Tables for reference year 1985 (8 countries)
C: Included in Eurostat Comparative Tables for reference year 1988 (12 countries)
D: Included in Eurostat Comparative Tables for reference year 1994 (18 countries including Iceland,
Norway and Switzerland)
E: Included in Eurostat Comparative Tables for reference year 1999 (16? Countries including Norway, ?
F: To be included in Eurostat Comparative Tables for reference year 2005
Keeping the 5-year cycle, Eurostat proposes to collect the next round of HBS data for the reference year 2005. With the same speed of data treatment as for HBS 1999 the results can be published at the end of 2007.
Eurostat proposes to collect HBS data for 2005.
2.2Presentation of annual results
The use of the European HBS is severely hampered by the fact that the data is only collected about every five years. Still the development is towards annual surveys, but for some MS there are no plans to increase the frequency. There is a certain demand for annual European results and Eurostat would be able to fund partially a project to develop estimates of annual results for the EU if some MS were interested to form a group to perform this work.
Given the present political demand for social statistics in other areas Eurostat cannot start collection of annual micro data. Therefore Eurostat proposes to investigate the possibility of collecting a limited set of standard tables to be put in the Eurostat New Cronos database. The use of the COICOP classification (or very similar) in all MS facilitates this task. Possibly there is need for the selection of a few harmonised background variables.
Eurostat proposes to develop an annual collection of a limited set of tables.
3Contents of the surveys
3.1Development of the COICOP-HBS
The COICOP-HBS was revised just before the latest Working Group meeting in April 1999, and the HBS1999 is presented according to the COICOP-HBS version 1999. At present there are no plans for the next revision of the COICOP. Nevertheless it is interesting to develop the details of the COICOP somewhat further.
The interest within the Commission lies within the following areas:
DG AGRI
- more details of price indices for food items
ESTAT - HCPI and Tourism statistics
- distinction between domestic and foreign consumption expenditure
DG SANCO
- food statistics for food safety
- electronic commerce
- cross-border trade
- details of financial services: consumer credit, mortgage credit, means of payment etc
- over-indebtedness
- package travel
ESTAT
- more details of cultural expenditure
- more details on private educational expenditure
All MS used the COICOP-HBS for the 1999 survey or are presently changing their classification into COICOP. Since there are national demands for greater detail, most MS have developed one or more additional levels, see below for some examples and some difficulties.
Function 1 Food and non-alcoholic beverages
- Fast food and frozen dishes (pizzas) are treated differently
- Food in restaurants and cafés is sometime subdivided according to type of food and sometimes according to place (domestic or not) and context (holiday or not) of consumption
- Fresh fruit and vegetables are treated differently compared to frozen fruit and vegetables
- No detail at all, not even split between food and drink
- Take away food and beverages are included
Function 4Housing
- Normal residence and free time residence are separated
Function 5Furnishing, household equipment etc.
- Amenities, durables – owner ship and purchases
The reference period for large purchase items (normally single items like cars, washing machines etc.) is usually one year. There are now Eurostat tables showing the percentage of households with a purchase during the reference year and the expenditure for that purchase. Indirectly this gives an estimate of how often purchases are made and the price, if it is a single item. Belgium, Germany, Spain and Italy have a reference period of three months for durables so for these countries the reliability is less and data have not been included.
Function 6Health and education
- Detailed information about number of visits to the doctor etc are collected
Function 7Transport
- Fuel by country (domestic or not)
- Travel according to geographic area and not means of transport
Function 9Recreation and culture
- All holiday travel and hotel expenditure or even all vacation expenditure is included in package travel
- Package travel by mode of transport
Function 11Restaurants and hotels
- Accommodation according to type and country (domestic or not)
The list of examples is far from complete. For the development of the COICOP-HBS a thorough investigation of the national practises would be done. Furthermore there should be an investigation to assess more clearly the needs of the Commission.
Eurostat will initiate a work to develop the detailed levels of the COICOP and to improve the descriptions.
3.1Development of variables on housing, amenities and durables
Statistics on housing, amenities and durables have been collected both in the HBS and in the European Community Household Panel (ECHP). Since the ECHP has provided annual data and the HBS only 5-yearly data, Eurostat has chosen the ECHP as the community source for housing statistics for example in New Cronos. The timeliness of the ECHP data on the other hand has not been acceptable (now the 1997 data is available). In the new instrument for Statistics on Income and Living Conditions (SILC – to start in 2003) some housing variables will be included. The number of housing variables (a draft list of variables is not ready but may be available for the meeting) will be reduced to a minimum. In order to meet the demand for more detailed statistics on housing a module might be introduced in the HBS with the aim to measure both physical variables and expenditure.
The list of HBS variables needs a revision in order to take new objects into account.
Eurostat will initiate a work to develop the recommendations for housing variables.
3.3Other issues
The HBS is a quite extensive and complicated survey. Still in some MS the HBS covers also additional subject areas such as food consumption, income, debts, type of outlet. Even time use is included in some MS. This makes the survey very large and the non-response rate risks to increase. A method to keep the survey at a reasonable size could be to introduce modules covering specific subject areas. The topics that could be suitable for an in-depth investigation are for example:
Food consumption
Since several Framework programmes the Commission funds food research covering aspects like food production, food safety, health and nutrition. The DAFNE (Data Food Networking project), funded by the Commission, has shown that HBS data can be used to obtain useful comparable data on food consumption. According the website the
“DAFNE databank currently comprises data on ten European countries, namely Belgium, Germany, Greece, Ireland, Luxembourg, Norway, Hungary, Poland, Spain and the United Kingdom, and allows:
- Inter-country comparisons of daily individual food availability, at different levels of detail, ranging from 45 analytical to 12 aggregated main food groups and
- Comparisons of the eating behaviour of various population segments. Three socio-economic parameters with important public health implications are used for within and between countries comparisons: the degree of urbanisation of the area where the household is situated, the education and the occupation of household head.
Recently, the DAFNE databank has been enriched by data from three European countries: France, Italy and Portugal.”
Eurostat has not been involved in this project
Assets and debts
The question of over-indebtedness is on the political agenda concerning social exclusion and consumer protection. The new Statistics on Income and Living Conditions (SILC) will cover some variables on housing loans and other loans but the information will not be detailed.
Children
“How much do children cost?” Eurostat has received this question several times and we have no statistics to describe the monetary expenditure. Some MS have introduced this aspect in the HBS
Domestic/foreign expenditure - tourism
Fairly detailed tourism expenditure are requested in the Tourism Statistics Directive and some member states have included this in the details of their national COICOP.
Type of outlet and means of payment
Some MS have included these variables in the HBS and for future European comparisons it would be valuable to investigate the variables and propose harmonised definitions.
Financial services
Financial services, for example bank charges and insurance services are difficult to measure. Comparable household statistics in these sectors have been demanded both by DG SANCO and the interest organisations.
After agreement of the Working Party Eurostat would start the work on concepts and definitions for some of the above subject areas.
4Data collection from the candidate countries
If the political demand for statistics on low income and social exclusion implies collection of income data from Household Budget Surveys in the candidate countries, Eurostat will collect both income and expenditure data from these countries. The draft project plan for income statistics includes data collection and setting up a Eurostat data base in 2003.
1