/ EUROPEAN COMMISSION
EUROSTAT
Directorate E: Social and regional statistics and geographical information system
Unit E-2: Living conditions /

Doc. E2/HBS/134/01

Working Group

HOUSEHOLD BUDGET SURVEYS

01-02 OCTOBER 2001

Eurostat-Luxembourg, Bech Building – Room Ampère

Point IV/6 of the agenda

Quality of the HBS in Phare Countries - Report

Elita Pavlovska (HBS team, Unit E2), December 1999-February 2000

REPORT

1. COMPARATIVE ANALYSIS OF THE HBS METHODOLOGY

IN THE PHARE COUNTRIES

The description is based on the methodological information about Household Budget Surveys (HBS) of 11 Phare countries (Albania, Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia). For the FYR of Macedonia, there is no any wider and recent information concerning the methodology of its HBS at our disposal.

Frequency: In a majority of the Phare countries, Household Budget Survey (HBS) is a continuous (annual) survey, in the sense that the fieldwork takes place on a continuous basis throughout the year. Most of the countries produce quarterly results, as well. Only in Albania, HBS is not annual: the last survey was carried out in 1994, and it covered only the capital – Tirana. (A new (continuous) survey started in December 1999.) Slovenia started a new HBS in 1997. The statistical office of Slovenia has published some survey results on 1997 and 1998. Then they are going to publish cumulated results on three years together (1997, 1998 and 1999) and to continue to do it according to the “moving principle” in order to improve the reliability of the data.

Latest reference year: The latest reference year of HBS for which data is available in the majority of the Phare countries is 1998, only for Slovenia and the Republic of Macedonia it is 1997, and for Albania – 1994.

Sample size: The range of sample size for HBS in the Phare countries varies from 1,050 in the Republic of Macedonia to 36,072 in Romania. Mostly, the sample size depends on the size of the total population of the country.

Sample design: The majority of the countries use random probability sampling (one- or two-stage, or combined). In the Czech Republic and Slovak Republic – quota sampling has been used.

Sample representativeness: The countries reported under-representativeness of the following categories:

  • Households with high income (Czech Republic, Hungary, Latvia, Lithuania, Slovak Republic);
  • Households of very young/very old persons (Hungary);
  • One-person households (Latvia, Poland);
  • Self-employed (Hungary, Latvia, Poland);
  • Households of economically active persons (Latvia);
  • Urban households: mainly in the capital (Hungary);
  • Farmers (Poland);
  • Households with the source of income other than wages and salaries (Poland);
  • Households living in private houses (Romania);
  • Homeless people (Estonia, Hungary, Latvia, Lithuania, Slovak Republic, Slovenia).

Sample frame: As a sampling frame Phare countries use (micro-)censuses, lists of dwellings (addresses), population registers or some other kinds of population databases. Romania uses master sample: a randomly selected area sample of the updated Census of 1992. For some countries, for example, Lithuania, there is a problem of quick up-dating of the population register.

Unit of measurement: In all the Phare countries with the exception of Albania, the unit of measurement is a household. In Albania, it is a dwelling.

Response rate and its calculation method: Response rate for HBS varies from one country to another. The lowest rate of response for the latest survey was in Bulgaria (59%), Estonia (53.5%) and Hungary (57.6%).

Table 1: RESPONSE RATE IN HBS IN PHARE COUNTRIES, in per cent

AL / BG / CZ1 / EE / HU / LT / LV / PL / RO / SI / SK1
~92 / 59.0 / - / 53.5 / 57.6 / 87.3 / 78.3 / 65.7 / 89.7 / 81.7 / -

1 Response rate is not calculated (quota sampling).

The response rate in these countries is lower than the recommended minimum (60-65%). In Bulgaria, the response rate is calculated as the number of initially selected households participated in the survey divided by the total number of selected households. In Estonia, the response rate is calculated as the number of households that answered to all the questions in the “Household Picture” and that filled in the “Diary Book of Food Expenditure” and the “Diary Book of Income, Taxes and Expenditure”, divided by the number of originally selected households (the frame error being excluded). Substitution for refusals is not allowed in Estonia. In the other Phare countries, response rates were between 65.7 and 92% (see the table 1). In Albania and Hungary, the response rate is calculated as the number of households – data suppliers (including substitution) divided by the number of originally selected households. In Hungary, so-called primary response rate (before the substitution) is calculated, as well, and it was 51.5% in 1998. In Latvia and Poland, the response rate is calculated as a percentage of complete response cases in the total number of households taken both from the basic and reserve list after subtracting the number of persons (households) excluded from survey due to imperfections and errors in the population register. In Lithuania, the response rate is calculated as a percentage of complete response cases from the total number of selected households.

Substitution: Refusals are replaced in Bulgaria, Hungary, Latvia, Poland and the Slovak Republic. In Bulgaria, each randomly selected household that is not willing to participate in the survey is replaced by another household with the same number of members and from the same cluster as the first one. Similarly, in the Hungarian survey, systematic substitution (each household has another household with similar demographic characteristics) is applied. Latvia and Poland use sequential substitution from the reserve-sampling list that is formed exactly in the same way as the basic sampling list, thus reducing systematic sampling error. In the Slovak Republic, each household not willing to continue recording is substituted by a household having the same quota criteria.

MAIN CONCEPTS AND DEFINITIONS

Coverage: In all Phare countries, HBS is confined to the population living in private households. Collective or institutional households are excluded. In the Czech Republic, in addition, the following households are not included in the sample (quota sample):

  • households with a pensioner being head of household and other economically active members;
  • households for which one of the members is away for a considerable period of time and the household is not willing to keep a record of his/her expenses;
  • households whose heads are people neither economically active nor pensioners (e.g. students).

In several countries (Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovenia), foreigners are excluded. In Albania, only private households living in the capital were covered by the HBS conducted in 1994.

Definition of the household: Eurostat recommends that the definition of household for the purpose of HBS is based on the two following criteria: co-residence (living together in the same dwelling unit) and sharing expenditure including joint provision of essentials of living. In all the Phare countries having provided methodological description to Eurostat, the two mentioned criteria are satisfied, moreover, in the majority of the countries, the definition of household requests also for joint income management, not only expenditure (table 2). In Estonia, in addition to all the three criteria, the forth one exists: necessity for persons living together to consider one another to be members of one household. In the Slovak Republic, the definition of household requires only sharing the dwelling and common expenditure for food and accommodation.

Table 2: DEFINITION OF HOUSEHOLD

Households defined as persons sharing:
accommodation / expenditure / income
Albania / X / X / X
Bulgaria / X / X / X
Czech Republic / X / X / X
Estonia / X / X / X
Hungary / X / X / X
Latvia / X / X / X
Lithuania / X / X1 / X
Poland / X / X
Romania / X / X / X
Slovak Republic / X / X
Slovenia / X / X / X

1 Common meals are also one of the necessary condition.

Different categories of persons are treated differently from country to country. It concerns, for example, resident employees, lodgers, boarders and other unrelated persons living at the same address, and certain categories of the persons who are absent from the household for a long time (pupils, students, persons in military service, persons in hospitals). Inclusion/exclusion of the mentioned categories in/from the household will affect the average household size and composition, as well as the coverage achieved in the survey. For example, only in the Slovak Republic and Slovenia, resident employees and domestic servants living in the household are considered to be household members (see table 1 in the annex). Thus, in several countries, a considerable number of young people (mostly working students and other young people that do not maintain economical links with their parents household) are not included in the survey frame causing under-coverage of certain social categories. Persons who are normally long-term absent but are present during recording period, are considered as household members only in the Albanian, Estonian, Polish, Romanian and Slovak HBS. In Poland and the Slovak Republic, long term absent persons have to maintain an economic link with the household (to transfer their income to the household) to be considered as its members. Visitors are not included in the household in any of the Phare countries. Residents being temporarily away are included among household members in all Phare countries, but mostly on different conditions (see table 1 in the annex). For example, in the Czech Republic, residents temporarily away are considered as household members, if they keep recording of their income and expenditure during their absence. In Hungary, Latvia, Lithuania, Romania, Slovak Republic and Slovenia, the persons temporarily away have to maintain an economic link with the household in order to be considered as a member. Concerning long term absence, students and boarding school pupils are included in the household in all Phare countries with the exception of Lithuania and Poland, but several countries have specific conditions for considering this category of person as household member. In the Czech Republic, they have to keep recording their income and expenditure to be considered as household members, but, in Estonia, Hungary, Latvia and the Slovak Republic, they have to maintain regular economic ties with the household and to have certain material dependency from it. In addition, in Hungary and Latvia, students are not included in the household of their parents, if they earn money and have their own budget. Hospitalised persons are considered as household members in all the Phare countries, with the exception of Estonia. Thus, one can see from the table in the appendix that Albania has the least strict rules for excluding certain categories of residents from the household composition, and, therefore, it has broader coverage (in some cases maybe even over-coverage) of the population in the HBS. Lithuania, in its turn, has rather strict rules: only normal residents related to head/spouse, residents temporarily away, but financially supporting their families, and persons hospitalised for a maximum of one month and receiving food from his/her household are considered to be household members.

Reference person: Eurostat recommends that the assignment of “the reference person should be based on objective criteria such as contribution to household income; by preference the person to be chosen should be the one contributing most to the total income or budget of the household”.[1] In the majority of the Phare countries (the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic), a reference person is not defined in the HBS. Instead, they use the term “head of the household” who may be the person with the highest income, the oldest economically active man in the household, or a person defined by the household itself. Only 5 of 10 countries that have sent the information on the HBS methodology (Estonia, Latvia, Lithuania, Poland, Slovenia) define the main breadwinner or the person with the highest income in the household as the reference person (head of household). Moreover, in Estonia and Latvia, it is the long-term highest contribution to the household budget. In Albania and Hungary, the oldesteconomically active man is the reference person of the household. In the Czech Republic, in complete families, the reference person (head of the household) is always a man. In incomplete families, the economically active parent is usually regarded as the head of household. In non-family households, the person with the highest income is considered to be the head of household. In general, economically active persons have the priority. In the Slovak Republic, in households of complete families (with both of the parents), it is the man, in single-parent household: the parent, and in non-family household: the person chosen by the household itself. In Bulgaria and Romania, the household itself declares the reference person in all types of households.

In most of the Phare countries, like in the Member States of the EU, the child-adult definition can be adapted depending on the purpose of analysis. For example, in Slovenia, children are considered to be persons under the age of 14 years (for filling in the personal questionnaire), but some questions in the questionnaire relate only persons aged 15 years and more. In Albania, children are considered to be persons under the age of 15 years, in Estonia and Poland: under the age of 16, in Bulgaria, Latvia and Lithuania: under the age of 18, in Hungary: under the age of 20 years (if attending school). Czech Republic and Poland operate with wider limits concerning this definition according to the economic dependency of the person to the household. In the Czech Republic, a person up to 26 years is considered to be a (dependent) child, if (s)he after finishing the secondary school methodically prepares for her/his future job (carries out studies, apprenticeship), so is a disabled person of this age who is not able neither to work nor to study and does not receive full invalidity pension. In Poland, dependant children are considered to be persons under the age of 16 years, or under the age of 25 years, if they do not have their own source of maintenance and if they are not married. In the Slovak Republic, children are considered to be persons under the age of 15 years (under the age of 28, if they are students).

SURVEY CONTENT

Consumption expenditure approach: The measure recommended by Eurostat for the study of living standards is that of final consumption expenditure that “corresponds to the expenditure effected by households in order to consume goods and services. This includes, in additions to purchases effected in monetary form, the estimated value of certain goods and services, e.g. the value of internal production, the benefits in kind and the imputed rents for certain categories of households. On the other hand, investments effected by the households (e.g. purchase of a house, major works on housing), direct duties and taxes paid to the various administrations, and savings are excluded from this concept.”[2]

All the Phare countries record consumption of own production for all households. In some of the countries, households themselves evaluate consumption, in others – evaluation is done a posteriori using average price for the quantity, but in all the countries, retail price is used for the evaluation. In all Phare countries, with the exception of Albania, benefits in kind received by employees and from self-employment are registered. However, in none of these countries, imputed rent is estimated and included in the consumption expenditure. Only Slovenia has made the first steps in the estimation of imputed rent but it does not include the obtained value in household consumption due to the low reliability of the results. Thus, Phare countries have only partially adopted the final consumption approach in their HBS.

Household diaries and interviews: All Phare countries that replied on the methodological questions use intensive household diaries for recording daily expenditure. Albania, Bulgaria, Latvia, Lithuania and Slovenia use 14 or 15-day intensive recording period, the Czech Republic, Hungary, Poland, Romania and the Slovak Republic – one-month recording period. In Estonia, households fill in the Diary Book of Food Expenditure during half a month, and in the Diary Book of Income, Taxes and Expenditure: during one month.Only Slovenia uses individual expenditure recording diary. In Lithuania, household members have possibility to record their expenditure into the so-called pocket diaries, but, at the end of the survey month, all records from these diaries are re-written into the household diary.

The majority of the countries for recording periodical and exceptional expenditure use the diary with a recording period equal to one month. In Albania and Slovenia, information on periodical and exceptional expenditure, purchasing of durable goods and housing is asked by interviewer and is recorded in the questionnaire. In Latvia, some questions concerning durable goods are asked directly by the interviewer. In Poland, the questions on housing are asked by the interviewer at the beginning of the month. An additional interview is conducted after three months, and it contains questions about rare expenses and durable goods. In the Czech Republic and Hungary, retrospective interviews (on 12 months) are carried out containing questions on dwelling, household’s durable goods, in Hungary – also on personal and family income and on exceptional expenditure such as construction purchase of real estate and other large expenditure.

Answers on the questions about household’s or individual consumption of benefits in kind and about income are mostly recorded in diaries with the reference period equal to one month. In Latvia and Lithuania, questions on personal income of the household members, income of the household as a whole and financial transfers are asked directly by the interviewer. In Hungary, in January and at the end of the first quarter of the year, household members are asked about their personal income, and on family income in general - retrospectively for the whole year. In Slovenia, all questions concerning household income are asked by the means of interview.

Questions on household/individual characteristics are asked by the interviewer and recorded in the questionnaire during the first interview. In Estonia, if the household is surveyed for the second or third time, the interviewer records also the changes in the household that have taken place during last three months.