Declarations of content - Declaration

0 Administrative Information about the Statistical Product


0.1 Name
Household Budget Survey

0.2 Subject Area
Incomes, consumption and prices

0.3 Responsible Authority, Office, Person, etc.
Prices and Consumption
Henrik Sejerbo Sørensen tel.: +45 3917 3662
e-mail: hss@dst.dk

0.4 Purpose and History
The purpose of the survey is to give information on the economic conditions of the private households - incomes, savings as well as consumption. Consequently, the survey can be characterized as a survey of living standards ands of general economic conditions.

The survey has been conducted at varying intervals since the late 19th century. From 1994 the survey was subjected to substantial methodological changes, and the survey is now conducted on an annual basis in a way that makes comparisons possible.

Similar surveys are conducted in most other countries.

Other statistical products give information on topics in the household budget survey, e.g.:

• The national accounts give information on private consumption - but the figures from the household budget survey are more detailed, and they can be grouped by household types.
• The income statistics give information on household income - but the income concept in the household budget survey is broader and covers also untaxed incomes.
• The general statistics on households, housing, etc. give information on households and on housing - but the household concept in the household budget survey differs slightly, as it is based the own definition of the household.

In general, it is recommended that the household budget survey is only used (based on a sample) in cases, where the information needed cannot be found in other (register-based) statistical products.

0.5 Users and Application
In addition to serving the general public interests the survey has broad groups of users:

• It is used internally at Statistics Denmark in compiling price indices and national accounts statistics.
• Government bodies use the survey for purposes of planning and for conducting analyses of the consequences of new legislation, etc.
• The survey is used for researches purposes within several fields.
• The survey can be used for marketing purposes, etc.
• Internationally, the survey is widely used. Especially, EUROSTAT is very active in enabling comparability of the survey results among the EU Member States.

0.6 Sources
The survey is based on a sample, and the number of participating households is approximately 2,500 from among 2.6 million households in Denmark. (estimated in the Household Budget Survey).

The data is obtained from 3 different sources:

The households have participated in a comprehensive interview. Here, they have been asked about regular outlays a year back, major expenses, the stock of durables kept by the households, the use of services within the health system, education and child care, and some incomes. The interviews are conducted with different households throughout the calendar year, so some data will concern the previous year.

The household has kept a detailed diary for two weeks, implying that the households write down every single expense in this period. The diary keeping is conducted in different households throughout the calendar year, so seasonally changes are taken into account.

The diary keeping has two main purposes:

• It ensures a total coverage of e.g. new products and rarely consumed products.
• It ensures coverage of more day-to-day consumption that cannot possibly be remembered a year back.

To ease the participation in the survey, questions regarding topics for which Statistics Denmark already have usable data, are left out. This regards data about income and taxes, dwellings, education and occupation.

The purpose is to draw a sample, which gives results that are good approximations for all private households in Denmark. At the same time, it is important that the interviewers conducting the data collection are the best qualified for this job.

It is not possible to draw a sample among households in accordance with the way in which they are defined in the survey. Instead, the sample is drawn from addresses. When the interviewer visits the address, it is decided whether one or more households are living at the address, or if the address is unoccupied or occupied by a shared household, such as an old-age home, which is to be excluded. In the case of more households at the address, all households have, as far as possible, participated.

Different addresses were visited at different times of the year, so that the whole year has been covered. The participating households were visited twice:

• At the first visit the household was defined, the instructions of the diary keeping were given and the first part of the interview was completed.
• At the second visit two weeks later the second part of the interview was completed and the interviewer received the completed diary from the household.

The interview has been carried out using a laptop. Consequently, it has been possible to do a number of checks at the spot, which is important for the quality of the survey.

The objective is to collect data from approximately 865 households a year. A sample of this size is too small to give a reasonable foundation for a detailed statistic, so instead data from 3 years in succession are aggregated into one sample. All expenses, incomes and so forth are recalculated to the price and volume level of the middle year. In the 2004 survey, data is collected in 2003, 2004 and 2005, and subsequently recalculated to the 2004 level.

0.7 Legal Authority to Collect Data
Data is collected according to section 1 of the Act on Statistics Denmark.

Participation in the survey is voluntary.

0.8 Response burden
No response burden has been estimated since participation in the survey is voluntary.

0.9 EU Regulation
No direct regulation exists - but the regulation on the harmonized index on prices indirectly presupposes that a kind of household budget survey is conducted.


1 Contents


1.1 Description of Contents
The survey gives detailed information on the economic conditions of the households. The following main topics are covered in an exhaustive and comprehensive way:

• Consumption
• Incomes
• Stock of durables
• The use of health-, education- and child care services
• Pension schemes
• Direct taxes
• Taxes on imports and production
• Indirect transfers from the public - i.e. the transfers, which the household receives without paying the total costs for the services.

This information can then in a great variety of ways be combined with background information on the households, among this information is:

• Household size and composition
• Household income
• Housing conditions
• Level of education
• Geography
• Etc.

1.2 Statistical Concepts

Units and population:

The household budget survey examines the economic conditions of private households. Consequently, consumption is recorded in value, not in volume. However figures are compiled in both annual and fixed prices.

The survey examines the actual economic conditions of the households. Questions about attitude or opinion are not collected.

A private household is defined as an economic unit: A group of persons living together and having a high degree of common economy - that is sharing incomes and expenses. There are approximately 2.6 million private households in Denmark (estimated by the Household Budget Survey).

Persons living in different kinds of shared households (prisons, hospitals, some institutions etc.) are excluded, since it is typically impossible to distinguish the private economy from the shared economy. For some of the survey variables, it is important to have this exclusion in mind, e.g. for analyses of the use of hospital services (as more permanent hospitalized persons are not covered).

In principle, the survey covers 99 percentage of the total population.

It is the private economy that is examined. Goods consumed in a production process (plant and machinery etc.) and other company-related expenses are not covered.

The household concept in the Household Budget Surveys differs from other concepts used by Statistics Denmark. In the register-based statistics, households are either defined as persons living at the same address, or as persons connected by different bonds (marriage, common children etc.). In the Household Budget Survey, the definition is made by the involved household itself in co-operation with the interviewer.

In most cases, there is no difference between the different definitions, but in some cases a lodger will be included in the main household in the register-based statistics, while the lodger who has a separate economy constitutes a separate household in the Household Budget Survey. In the typical statistics from the survey, no distinction is made between married and non-married couples, and no distinction is made between couples of different or same sex. Furthermore, it is not important if a person is registered at a specific address at the national registration office. It is the actual situation when the interviewer visits the household that is important.

While the data in the survey is normally per household, it will in many cases be relevant to look at the data per person or by equivalence unit.

E.g. it is not meaningful to compare an average household whose yearly income is over DKK 800,000 and whose total consumption amounts to DKK 429,000, while an average household whose yearly income is under DKK 150,000 and whose total consumption only amounts to DKK 113,000. This is due to the fact that the first household type consists of 3.2 persons, while the latter consists of 1.0. Consequently, more people share income and consumption in households with higher income.

For analytical purposes it can therefore be relevant to look at the consumption per person in this case the consumption in households with higher income will be DKK 134,000 per person, while the household with lower income uses DKK 113,000 per person for consumption.

In a welfare analysis conducted on the basis of consumption per person, it is indirectly assumed that all persons have the same needs, regardless of the household type to which they belong.

To make a better comparison between households of different size and composition, consumption etc. can alternatively be calculated by equivalence unit. The reason is that households consisting of several persons have certain large scale advantages. A household with several persons are able to share a part of the dwelling area and installat tions, it may make large scale shopping cheaper, every person in the household does not need a car etc. To obtain a similar welfare level, a household with 2 persons does not need to have a twice as large consumption as a household with 1 person. Furthermore, it is implied that children do not need to have the same consumption level as adults, to obtain the same welfare level.

To compare the welfare level between different household types, an equivalence scale can be used. This type of scale is designed to recalculate the welfare effect of consumption or income of different households. The main unit is the adult equivalent - that is the consumption in a household consisting of precisely 1 adult. There is no unambiguous way to define such a scale, so on the basis of international recommendations Statistics Denmark has decided to define the equivalence unit in the following way:

The first person over 14 years old counts for 1, other persons over 14 years count for 0.5 and children 14 years old or lower count for 0.3. A household consisting of 2 adults and 2 children therefore consists of 2.1 equivalence units. This scale is called the modified OECD-scale.

Re-estimated using this scale a household with an income of 800,000 consists of 1.9 equivalence units, while a household with an income of up to DKK 150,000 consists of 1.0 equivalence unit. The consumption per equivalence unit is therefore DKK 226,000 in the first household type and DKK 113,000 in the latter.

This way of defining the scale is subjected to discussion. It may be claimed that the scale should be graduated even more by the age of the person, that the large scale effect should be graduated and so forth.

While consumption is collected by household, other parts of the survey are collected by person. This applies to incomes, taxes, data about pension schemes and the use of public services etc.

Variables:

Wages and salaries etc. consists of wages and salaries before tax is paid, employers and employees payments to pension schemes, including ATP (Danish Labour Market Supplementary Pension Scheme), as well as salaries in non-monetary forms e.g. paid for car, dwelling etc.

Entrepreneurial income etc. consists of profit from self-employed occupations. Investment income regarding the business is included, while interest expenses are excluded. The item can sometimes contain values below zero either because the business as such has a deficit or because the interest expenses have been large. Furthermore, self-employed person consumption of own goods from production or shop (e.g. farmers or grocer) is included, as well as income from undeclared work.

Property income contains income from private investments and stocks etc. Furthermore, the imputed rent for owner-occupied dwellings and secondary residence are included. On the other hand, increase in value (either realized or not) of capital goods such as securities and real estate is not included.

Private transfer income contains received presents and winnings, payments from insurances etc. as well as payments from private pension schemes (including ATP) and received child maintenance and alimony.

Public transfer income contains payments from public pensions, unemployment benefit and other unemployment fund payments, sickness benefit, cash benefit, study benefit, housing benefit as well as child benefit.

Other incomes and balancing contains several minor types of income, as well as balancing as a result of difference in periods between different types of data. The value may sometimes be below zero.

These incomes add up to the gross income.

Besides these types of more current incomes, the household also may receive capital transfers, such as inheritance, payme ents from the Danish Employees Capital Fund, one-time payments from ATP etc.

The sum of the gross income and capital transfers constitutes total income.

From the total income income taxes etc. (direct taxes) are paid, which include ordinary income taxes, labour market contributions, taxes regarding inheritance, housing and presents as well as payments to unemployment insurance.

Furthermore, private interest expenses etc. are paid, in which especially interests regarding housing are of importance. Also other compulsory transfers are included, such as paid child maintenance, alimony, compensation etc.

The total income minus income taxes and private interest expenses constitutes the disposable income.

Adding the net sum from received capital pension to the disposable income equals the central concept disposable amount.

The disposable amount can be used to pay different fines that is a payment, which is similar to taxes.

It can also be used for presents and charity towards other households or charity organizations.

It can be used to pay for subscriptions to associations (non-profit organizations regarding households), including labour unions, sport associations etc.

The household can also use there disposable amount as savings, which can be divided into four types:

• Payment for pensions schemes and ATP (Danish Labour Market Supplementary Pension Scheme). This includes all types of pension schemes, private as well as types organized through an affiliation at the labour market.
• Payment for private life insurances etc.
• Value of extension and rebuilding etc. of the dwelling or secondary residence as well as expenses regarding building of a new dwelling considered to be investments regarding the dwelling. According to international recommendations, expenses regarding buying or selling of the dwelling are included (e.g. the real estate agents fee). However, the buying or selling price in itself is excluded.
• Other kinds of savings, which are calculated as a residual, and cover all other forms of positive or negative savings, including repayment on loans on owner-occupied dwellings, consumer credits, student loans, savings or loans in banks etc.

Please note that in spite of payments to pension schemes, they are included as positive savings, payments from pension schemes are regarded as income and not as negative savings. The reason is that for the individual person, there is no direct relationship between the contribution to and the payments from the pension scheme, because it depends, among other things, on the expectancy of life for a person.

The remaining part of the income is used as consumption.

• The major part is paid consumption of goods and services. Note that goods bought by instalment are included as the total cash price (excluding credit costs). Repayments are, on the other hand, not included as consumption but as saving. Private sale of personal property is included as negative consumption. This is especially important concerning cars, where it is common to pay a part of the new cars price by means of the sales price of the old car.
• The imputed rent for owner-occupied dwellings and secondary residence are included as consumption.
• Furthermore, also types of non-monetary incomes are included (paid for car, telephone as well as self-employed persons consumption of own goods).

As supplement to these main aggregates, also data about indirect contributions from the authorities, which are indirectly received by the households when using public services as child care, education and health care, which is free of charge or is subject to a price reduction, are available.

Fu urthermore, data about production- and import taxes (indirect taxes) paid by the households are available, i.e. taxes and duties imposed especially on the consumption, but also extension and rebuilding of the dwelling, such as VAT on building materials etc. and stamp duty regarding buying and selling of dwellings.

When including the indirect contributions and production- and import taxes, it is possible to monitor the economic transactions between the households and the authorities:

• To the household from the authorities: Direct transfers, i.e. public transfer incomes + indirect contributions.
• From the households to the authorities: Income taxes + production- and import taxes.

Especially the indirect contributions are difficult to interpret. It makes no sense to claim that at person being hospitalized over a longer period, and therefore having received a large indirect contribution, has a welfare gain in comparison with a healthy person. Furthermore, it is worth noticing that only some but essential types of contributions are included. A major drawback is that it has not yet been possible to include public contributions towards medicine.

Statistical measurements:

The typical statistics from the survey are given as an average in DKK per household per year.

Groupings:

Consumption as well as incomes can be grouped at different levels of detail. The consumption can for instance be classified to between 11 and 1,200 groups.

Fixed prices
Figures are compiled en fixed prices. All figures are compiled in 2005 prices.


2 Time


2.1 Reference Period
The data is collected over a period of 3 years, so the real reference period is this 3-year period.

To facilitate the use of the data, all information is recalculated in the best possible way so that it covers the year in the middle, implying that for the practical use, the reference period is one calendar year (for flow variables, such as consumption and income, it can be perceived as the calendar year, and for stock variables, such as household configuration and dwelling situation, it can be perceived as the end of the calendar year).

2.2 Date of Publication
The statistics are updated tvice yearly, where approximately one third of the households are new, while two thirds was also included in the previous survey. Preliminary figures are published approximately 6 month after the last data collection. Final figures approximately a year.

2.3 Punctuality
The statistics are usually published without delay in relation to the scheduled date.

2.4 Frequency
Two yearly updated version is published. One preliminary and one final.


3 Accuracy


3.1 Overall accuracy
A survey like the Household Budget Survey is subject to a number of inaccuracies. Most errors and shortcomings are not of a kind that can be measured, and it is therefore not possible to measure the total inaccuracy in the survey.

The sample-related coefficient of variance for total consumption per household is estimated at less than 1 pct. The coefficient of variance is estimated on the assumption that the sample is simple and random.

The total inaccuracy of which the sample related coefficient of variance is only a part can, as noted, not be measured, because it is not possible to measure the other types of errors. What other kinds of errors should be taken into account are described in the next section.

In general, the inaccuracy is higher, the more detailed level data are broken down to and the fewer households on which the average is based.

3.2 Sources of inaccuracy

Coverage:

The survey is based on a sample, where the final number of households in the survey is approximately 2,500 from the total of
approximately 2.6 million private households in Denmark (estimated in the Household Budget Survey).

Persons living in different kinds of shared households (prisons, hospitals and other institutions) are not included, because it is difficult to separate the private economy from the shared economy. In principle, the survey covers 99 percentage of the total population.

Sample:

The latest survey consists of data from approximately 2,500 private households, randomly chosen .

The response rate is approximately 50 pct. A response rate of this size is considered to be fair especially when considering the extensive efforts related to participation for the responding households.

Also in comparison with similar surveys conducted in other countries, the response rate is satisfactory.

Participation in the survey is of course voluntary. Statistics Denmark is not able to give any real payment, but the participating households receive a gift and participate in a lottery in return for their participation.

Data collection:

The data sources are interviews, diaries and administrative data, where different kinds of inaccuracies are associated.

• The households lack of memory might influence the interview, because it can be difficult for the household to remember, if a specific expense was made 11 or 13 months ago.
• In keeping the diary the household can forget to register certain expenses especially at the end of the two-week period. Analysis has shown that more expenses are registered in the first week than in the second. To adjust for this forgetfulness, all diary data from the second week are enumerated with 4 pct.
• The households lack of knowledge might influence the quality of the difficult questions about insurances.
• The households reluctance to give correct answers might cause a problem. For instance, it is possible that answers about undeclared work is faulty and therefore give underestimated values, because some households have been reluctant to answer correctly. The same problem might affect answers regarding consumption that is illegal (e.g. narcotics) or in some way is perceived as problematic (e.g. a large consumption of alcohol). In general though the impression is that the participating households are both willing to answer questions and to be honest. An imbalance rather occurs because households with something to hide do not participate in the survey.
• Errors and shortcomings in the administrative register data might cause a problem. For instance, it cannot generally be expected that the tax authorities always correct incorrect data if there is no tax-related consequence. Furthermore, register data are defective for households where a person has died and for households which were not assessed ready (typically with complicated income and tax) at the time the data was obtained by Statistics Denmark.

Non-response:

Non-response is unwanted for two reasons: It weakens the final output, because fewer households participate, which makes the inaccuracy greater. But more troublesome it adds bias to the results, when the non-response is unevenly distributed between the households.

The non-response is especially large among households with just one person or with 6 persons, among the elderly, among retired persons, among households with low income and in the metropolitan area and in municipalities with cities having between 40,000-99,999 inhabitants. Conversely, the non-response is small among households with 4 persons, among households where the oldest person is between 20-29 years old, among w wage earner households, among households with an income between DKK 500.000-799.000 and in municipalities with cities having between 10,000-19,999 inhabitants.

To partly neutralize the bias from the non-response and the sample, the values are weighted.

Household types where the non-response is large, and which are consequently underrepresented in the survey, are assigned a relatively large weight, and households which are overrepresented are assigned a relatively small weight.

In most tables, information about the estimated number of households in Denmark as well as the actual number in the survey is included. The latter information is important when assessing the inaccuracy, because a small number of households in a given group cause greater inaccuracy.

From the 2002-2004 survey a new method to calculate weights was implemented. Before the weight was calculated by post stratification, where the sample and the population were split into a number of identical strata, constructed from the combination of 6 characteristics: Household size and composition, income, the main income earners socioeconomic status, if the household owns or rents the dwelling and in what type of city area, the household lives. The principle is then that households consisting of 1 person, with an income of DKK 300.000, who is a wage earner, who rents his dwelling and who lives in the metropolitan area is assigned a weight, which is equal to the number of this particular type of household in the population.

The weights are calibrated using a regression estimate. The focus is on the individual characteristic in the relationship between sample and population. The advantage with the calibration method is that more characteristics can be included, than in the preveious method of post stratification. The following charachteristics are included: Household size and composition, income, the main income earners socioeconomic status, if the household owns or rents the dwelling, in what type of city area, the household live, level of education, gender, geography and other characteristics.

Processing of data:

An extensive checking and processing of the data is carried out comprising e.g.:

• For all amounts given by the households, it is checked if the size of the amount seems reasonable. For the interview-based data this is done at the visit to the households and corrections are made directly.
• For a number of data it is checked for the logical coherence.
• It is checked that there is a reasonable coherence between the income and the uses of income.

3.3 Measures on accuracy
See the description in part 3.1.

For the consumption the inaccuracy is greatest for goods seldom bought, for data from the diary rather than the interview, and for data from small subgroups of households rather than large.


4 Comparability


4.1 Comparability over Time
From the start of the 'new' survey in 1994 the data are highly comparable.
It should be noticed that the data to one year's update is collected in 3 years. When next year's update is computed it is done by changing the oldest year from last year's update with a new year. This means that in 2 different year's versions just 1/3 of the households are changed. When comparing data from different years it is therefore advised always to go at least 3 years back.

Comparison back in time is hampered, as the definitions and classifications have been widely changed.

4.2 Comparability with other Statistics
The classifications and definitions used are aimed to be as comparable as possible with the national accounts. But the methods, etc., in these two kinds of statistics are very different.

The classification of consumption is based on the international COICOP classification, which is also used in compiling price indices.

It is difficult to conduct comparisons with the general population statistics and other register-based statistics, as the household definition in the survey is different: In the survey the household definition is the economic unit, which is decided by the household members themselves, while in the general population statistics the household definitions are derived from the administrative registers-based information.

4.3 Coherence between provisional and final statistics
The coherence is estimated to bee good.


5 Accessibility


5.1 Forms of dissemination
General publication in

• Nyt fra Danmarks Statistik (News from Statistics Denmark)
• Statistical Yearbook www.dst.dk/en/Statistics/ofs/Publications/Yearbook.aspx
• www.statbank.dk/fu5, fu6 and fu7.

To this is added e different publications focusing on specific topics, e.g. education and consumption, geographic difference in consumptions patterns and the difference between households with high and low income.

EUROSTAT publishes data from the Danish surveys as well as from the other Member States about every 5th year. These EUROSTAT publications do not always use exactly the same definitions, etc., as in the national publications.

5.2 Basic material: Storage and usability
The data from the survey is stored as SAS-datasets on the PC-network.

On the basis of the micro data it is possible to make special tabulations, etc.

It is also possible for researchers to have access to the micro data under certain restrictive conditions.

5.3 Documentation
A detailed documentation is published in the book 'Forbrugsundersøgelsen. Metodebeskrivelse. Fra dataindsamling til offentliggørelse'. Statistics Denmark 1999. ("The Household Budget Survey. Methodological Description. From Data Collection to Publication").

A special paper (in Danish) on the possibilities of obtaining statistics on smaller geographic areas can be obtained by contacting the Section for the Household Budget Survey.

5.4 Other Information
Other information can be obtained by contacting the Section for the Household Budget Survey.


Supplementary documentation
No supplementary documentation is available