EUROSTAT
Directorate E: Social statistics
Unit E-2: Living conditions /
Doc. E2/HBS/135/2001/EN
Working Group
HOUSEHOLD BUDGET SURVEYS
05-06 MAY 2003
Eurostat-Luxembourg, Jean Monnet Building, Room M4
Point II of the agenda
Adoption of the Summary Recordsof the WP meeting 01-02 October 2001
1
/ EUROPEAN COMMISSIONEUROSTAT
Directorate E: Social and regional statistics and geographical information system
Unit E-2: Living conditions /
Doc. HBS/135/01/EN
Final draft: 12/12/2001
Working Group
HOUSEHOLD BUDGET SURVEYS
01-02 OCTOBER 2001
Eurostat-Luxembourg, Bech Building – Room Ampère
Draft summary records of the
WP meeting 1 - 2 October 2001
1
Introduction
§1. The chairman welcomed the participants of all NSIs, the Candidate countries, DG EMPL and DG SANCO. He presented the members of the HBS team: Mrs. Karin Winqvist (head of section), Mr. Antonio Puente (official in charge of HBS, replacing Mr. Harry Bierings), Mr. Nicholas Woerner and Mrs. Nassera Cheriguene (responsible of data bases) and Ms. Sara Jarvis (secretary). The objectives of the meeting were:
· To discuss the revision of methodology and COICOP-HBS;
· To discuss the harmonisation of a set of background variables;
· To present the results of HBS 1999 wave. The chairman remarked that final data were available for 11 countries and there was a significant progress speeding up the availability of results compared with the 1994 wave. He remarked the convenience of having yearly estimates;
· To discuss the future of HBS;
· To inform the Working Party about Statistics on Income and Living Conditions SILC (a new project launched in order to replace the European Community Household Panel, ECHP, in the near future).
· To inform the Working Party about some uses of HBS data, in particular a study on poverty, statistics on consumer protection and a survey on ICT.
I. Adoption of the agenda
§2. The chairman informed about some changes in the timing of the agenda and some practical information.
§3. The chairman also informed that the presentation of Household Consumption, initially scheduled for the seminar with Phare countries to be held on Wednesday, has been moved to Tuesday at 12:30.
§4. Mrs. Winqvist gave some explanations regarding the documents prepared for this Working Party. The most relevant were the following:
· Document HBS/124/01 on “Results” is based on preliminary results.
· A few documents could not be translated and they are only available in English.
· There is a document containing tables for the methodology. The Member State delegates were asked to revise or to complete some of these tables. She informed that information on table 9 of this document is not correct, so it should be taken out.
· Delegates were also asked to revise their corresponding survey profiles in order to be included in the new methodology.
§5. The agenda was adopted.
II. Approval of the summary records of the meeting of the Working Group on 19-20April 1999 in Luxembourg (doc. HBS/123/99)
§6. France made a correction on the recording of reimbursement of health expenditure in France. The recording is made before reimbursement and not after. This correction will be included in the final minutes.
III. Items submitted in the Working Group for discussion
IV. Items submitted in the Working Group for information
1999 Data Collection – Results (doc HBS/124/01)
§7. Eurostat presented the document, mentioning the following remarks:
· Presented data were an advance of a final document on results, not finished yet.
· These data belong to 13 countries, as data from 2 countries are not available yet.
· The objective of the presentation was to show some possible uses of the collected data at Community level without pretending to be either comprehensive or systematic.
§8. The attendants made some comments or placed some questions as follows:
Italy said that the Italian survey was completely re-designed in 1996. Therefore there is a break in the time series in 1996, and not only micro data before and after this date but also macro data cannot be directly compared. Regarding housing typology, single adult households usually mix very young people with old people living alone. So for certain analyses using the age of the household head could be preferable.
Finland provided three comments: What was presented is private consumption and, if we take into account the individual government consumption the picture made between the countries changes remarkably. The alcohol and tobacco consumption is reported in much smaller quantities then actually consumed by the households in Finland. In the 1998 survey (for reference year 1999) they have used the net expenditure concept instead of gross expenditure concept used until mid nineties.
Netherlands asked if there was still time to do corrections of the already presented data in order to improve comparability.
Eurostat mentioned that the publication on these results in preparation would be available by the end of this year. This publication will contain very few comparisons between 1994 and 1999 because there are comparability problems, not only for Italy and Finland but for other countries as well. Alcohol and tobacco is underestimated in most countries. Data of all countries is considered as final, except for Austria and Ireland (currently under revision) and France and Portugal (not available yet), unless serious mistakes are found. Comparability of data will be discussed later in the context of COICOP-HBS revision.
Concepts and definitions (doc HBS/125/01 for discussion)
§9. Eurostat presented the document containing the draft of a new methodology which is intended to achieve the following goals:
· Better integration of HBS within National Accounts;
· Wider harmonisation of background variables in conformity with what is being done by the countries at the moment, in order to improve comparability between countries;
At the end of the presentation, the speaker asked the following questions (in the field of social statistics) to the delegates
1. Is there a tendency in your country to make surveys more useful for integrative frameworks such as NA
2. If yes: Is this also the case for the HBS, or is the HBS seen as an independent source in your country?
3. For quite a number of countries, the gap between HBS and NA consumption expenditure grew between 1994 and 1999. How come?
4. Do you see possibilities in the medium term to integrate the HBS in the National Accounts by using the actual consumption expenditure concept? And to use: registers or modelling or the HBS questionnaire, if necessary to reach this goal?
§10. The country delegates commented these questions in a “tour-de-table”. The comments were as follows:
Belgium said that the National Statistical Institute no longer deals with the NA.
Denmark said that HBS and NA are two different statistics with different uses. They consider it a good idea to know what are the differences between HBS and NA and where they come from, but not necessarily to remove them.
Germany agreed with the Danish delegate that there are efforts to be made in order to facilitate an entire comparison of NA's and HBS's data. In Germany, HBS is used for the evaluation of certain NA estimators. Furthermore, there is a high interest in harmonising the characteristics and data of both statistics, in order to achieve the important objective of having coherent statistics. For this reason, discrepancies between HBS's and NA's results are monitored, and the causalities for those discrepancies are investigated.
Greece said that a consultation with his colleagues of NA is required to be able to answer these questions.
Spain tried to harmonise their definitions with the NA team in order to reach the household expenditure level. However, integration at micro-data level is difficult and their users want neither modelling nor subjective estimations.
France said that HBS is not used very much by the NA. HBS is carried out every 5 years and NA need annual sources. For France, it is difficult to do that kind of integration and it is not a priority for them. However, they would like to improve the survey, so it is easier for the NA people to use it. They are in contact with the NA in order to explain where the differences come from; but for the moment, the are not considering integration at all.
Iceland agreed with the Danish delegate as well.
Portugal is doing an annual survey and they know that their colleagues of NA are doing a household survey too. They expect to consult their colleagues of NA in order to harmonise the concepts and the definitions, so they can use the data of HBS.
Finland thanked Eurostat for taking the initiative of widening integration between NA and HBS. In Finland there is a tendency of wider co-operation between NA and HBS. Regarding the gap between NA and HBS, the Finnish delegate is not sure if it is increased or not. He sees possibilities in the medium term to integrate HBS in NA and modelling would be the best way to approach it.
Sweden said that integration between NA and HBS is not a priority for them. The gap between HBS and NA has been increased from 1994 to 1999 due to some methodological changes in the HBS. For example food is only covered as a lump sum and the measurement period has been reduced from four weeks to two weeks.
United Kingdom NA aims to use the best available source for each product or service; sometimes it is HBS, and sometimes it is another source. NA would like to use the HBS more. However, there are some disadvantages of HBS, such as the underestimation of the consumption expenditure of alcohol and tobacco and the sampling errors. UK is not aware of the increased gap between the measures of HBS and other sources.
Switzerland NA not only use HBS but it also uses other sources. Where possible, they would like to harmonise the concepts and definitions. With regard to the use of reciprocal results, they try to explain where the differences lie.
Irish HBS team co-operate with their colleagues of NA in the definition of concepts, questionnaire design, etc. The main disadvantage of HBS is that it is only carried out every 5 years. As the UK, they have problems with the underestimation of some products such as alcohol. As regards to the gaps between NA and HBS, they are not aware that there has been an increase from 1994 until 1999. Most of the other gaps come from the fact that some services such as housing, education and health come from the State, and they are not included in HBS.
Italy NA uses the HBS quite extensively. But in some cases they try to improve HBS estimations using modelling or external sources. Nevertheless, both statistics are quite related and this is why HBS has been redesigned in order to meet better the NA requests. However NA and HBS are not the same thing, and they cannot be fully integrated. Some things cannot be measured using information coming from households or cannot be estimated properly using a sample survey.
Luxembourg said that from 1998 NA changed the questions of HBS (or the way in which they are answered) in order to be more usable. The result has been a bit disappointing for the NA people and the use of HBS has been less than they expected. In the future the HBS will be much closer to the NA.
The Netherlands said that there is certainly a tendency in the Netherlands to make as many integrated frameworks as possible. HBS is seen as an independent source in the Netherlands. At this moment the survey is being redesigned completely. So 2000 will be the last regular survey and from 2003 there will be a totally new HBS in the Netherlands. The gap between HBS and NA in the Netherlands did not grow because they use NA as a source of reference. At this moment they don’t use many registers, but in the new survey this will change. Currently they are using their own nomenclature, but from 2003 COICOP will probably be used. The concept of imputed rent, as used in NA, may be also part of the HBS.
Norway said that they deliver results from HBS to NA and NA people are modelling if necessary. There are no plans to integrate the NA concepts in the HBS. They agree with Denmark and other countries that it is important to keep these statistics apart because of the differences between the micro and the macro level.
Austria said that they are in close co-operation with their NA colleagues and they have compared the results, and NA is using HBS data. But, the last (1999 and 2000) was an independent survey and the next will be in five years.
§11. The chairman concluded that
q Although not everybody answered all the questions, the general spirit for most countries was to co-ordinate HBS and NA but not necessarily to integrate them. This is due to several reasons: the different purpose and use of both statistics, some practical difficulties (i.e. HBS is not a yearly survey in some countries), the differences between micro and macro-data, etc. However, everybody was in favour of this co-ordination and eventually with explaining the differences of data with NA.
Employment Policy and need for statistics (for information)
§12. Mrs. Deasy, from DG EMPL and a user of HBS data, gave a background of the policy area of social exclusion within employment and social affairs. In summary, she said:
After the Lisbon Council, it was agreed that the level of poverty in the EU was unacceptable and that was to become a policy priority. The Nice Council set some new objectives. All MS should submit a “National Action Plan on Social Exclusion” every two years based on common indicators. These indicators are crucial because it is impossible to compare the progress and the evolution of performance and achievements in this area without having common ways of measuring. Establishing these common ways of measuring is both a technical and a political process.
At the end of the Nice Council it was agreed that there would be a Committee on Social Protection set up, including a subgroup on social indicators. The Commission role is to develop European databases and to prepare reports.