Eurostat, the statistical office of the European Union

F1: Social Indicators, methodology and development, relations with users

2920 Luxembourg LUXEMBOURG

Income, consumption and wealth (ICW) statistics are experimental statistics computed by Eurostat through the statistical matching of three data sources: the EU Statistics on Income and Living Conditions (EU-SILC), the Household Budget Survey (HBS) and the Household Finance and Consumption Survey (HFCS). These statistics enable us to observe at the same time the income that households receive, their expenditures and their accumulated wealth.

The annual collection of EU-SILC was launched in 2003 and is governed by Regulation 1700/2019 (previously: Regulation 1177/2003) of the European Parliament and of the Council. The EU-SILC collects cross-sectional and longitudinal information on income. HBS is a survey conducted every 5 years on the basis of an agreement between Eurostat, the Member States and EFTA countries. Data are collected using national questionnaires and, in most cases, expenditure diaries that respondents are asked to keep over a certain period of time. HFCS collects information on assets, liabilities, and to a limited extent income and consumption, of households. The survey is run by National Central Banks and coordinated by the European Central Bank.

This page focuses on the main issues of importance for the use and interpretation of ICW statistics. Information on the primary data sources can be found on the respective EU-SILC and HBS metadata pages and following the links provided in the sections 'related metadata' and 'annexes' below.

Experimental ICW statistics cover six topics: household economic resources, affordability of essential services, saving rates, poverty, household characteristics and taxation. Each topic contains several indicators with a number of different breakdowns, mainly by income quantile, by the age group of the household reference person, by household type, by the educational attainment level of the reference person, by the activity status of the reference person and by the degree of urbanization of the household. The indicators provide information on the joint distribution of income, consumption and wealth and the links between these three economic dimensions. They help to describe households' economic vulnerability and material well-being. They also help to explain the dynamics of wealth inequalities.

All indicators are to be understood to describe households, not persons. Breakdowns by age group, educational attainment level and activity status refer to the household reference person, which is the person with the highest income. The only exception are the tables icw_pov_01, icw_pov_10, icw_pov_11 and icw_pov_12 for which the income, consumption and wealth of households have been equivalised such that equal shares were attributed to each household member. Values in tables icw_aff are calculated for households reporting non-zero values only.

Note on table icw _res_01 and icw_res_02: The indicator “Households” [HH] in icw_res_01 shows the share of households in the selection, which hold the corresponding shares of total disposable income [INC_DISP], consumption expenditure [EXPN_CONS] and net wealth [WLTH_NET] of the entire population. In theory, turning two of the three dimensions [quant_inc, quant_expn, quant_wlth] to TOTAL and the third one to any quintile, should result into a share of 20% of households. Nevertheless, this share is often below or above 20% of the total population of households in the country. The reason for this is that our figures are based on sample surveys. This means that the share of households corresponds indeed to 20% of households in the sample, however when we multiply each household of the sample with its sampling weight, the resulting shares of households in the total population differ from the 20%. If, for example, we disregard the income and wealth of households in our sample, the first consumption quintile contains the 20% of households with lowest consumption in the sample. However, multiplying this selection of households with their corresponding sampling weights may result into a different share of the total population. The “Households” [HH] indicator indicates the real share of households in the population that make up the theoretical quintile.

ICW statistics are produced in accordance with the relevant international classification systems. The main classifications used are: the 'International Standard Classification of Education' ISCED 2011 for the level of education, the 'International Standard Classification of Occupations' ISCO (88(COM) and 08(COM) from 2011) for occupation and the European Classification of Individual Consumption according to Purpose (ECOICOP) 5-DIGIT 2013 for consumption expenditure of households.

Detailed information about all these nomenclatures may be found in RAMON, the Eurostat’s classification server.

Statistics have been published with the following breakdowns:

ECOICOP

The ‘affordability of essential goods and services’ (tables ‘icw_aff’) is broken down by ECOICOP classes. In addition to the standard classes, the following aggregates have been produced:

 Age classes

Educational attainment level

Activity status

The activity status is derived from the SILC variable PL031 “Self-defined current ‘main activity status’. It captures the person’s own perception of their main activity at present.

Household type

Eurostat has developed a common classification for use in data collection surveys including HBS and EU-SILC as well as the subsequent presentation of indicators relating to income, expenditure, etc. The classification is constructed by the number of adult members and the number of dependent children living with them.

Degree of urbanization

The degree of urbanization designates whether households live in rural, densely populated or intermediate areas:

Overburden status

A household is considered “overburdened” if its total housing costs (EU-SILC variable HH070) 'net' of housing allowances (HY070g) represent more than 40 % of its disposable income ('net' of housing allowances): (HH070*12-HY070g) > 0.4*(HY020-HY070g)

At risk of poverty

A household is considered ‘at risk of (income) poverty’ if its equivalised disposable income (after social transfer) is less than 60 % of the national median equivalised disposable income after social transfers.

Severe material deprivation

Severe material deprivation is defined as the enforced inability to pay for at least four out of nine predefined material items considered by most people to be desirable or even necessary to lead an adequate life.

Low work intensity

Households where the adults (those aged 18-59, but excluding students aged 18-24) worked a working time equal or less than 20 % of their total combined work-time potential during the previous year.

At risk of poverty or social exclusion (AROPE)

Households who are either at risk of poverty, or severely materially and socially deprived or living in a household with a very low work intensity. For 2010 and 2015 data the target of the Europe 2020 strategy is applied.

Low levels of expenditure

See section 3.4 below.

Not applicable.

Household

According to Regulation 1700/2019, a ‘private household’ means a person living alone or a group of persons who live together, providing oneself or themselves with the essentials of living. This means that persons living in a private household share their income and expenses.

Household reference person

Following the definition of the Canberra Group, the reference person of the household is the adult with the highest income.

See: UNECE. Canberra Group Handbook on Household Income Statistics. United Nations; 2011. Available from: http://www.lisdatacenter.org/books/the-canberra-group-expert-group-on-household-income-statistics-final-report-and-recommendations/

Income

Household disposable income (EU-SILC variable HY020) is established by adding up all monetary incomes received from any source by all members of the household (including all income from work, investment and social benefits) — plus income received at household level — and deducting taxes and social contributions paid. This definition excludes:

• imputed rent - i.e. money that one saves on full (market) rent by living in one's own accommodation or in accommodation rented at a price that is lower than the market rent,
• non monetary income components, in particular value of goods produced for own consumption, social transfers in kind and non-cash employee income except company cars.

An exception has been made for Romania 2015: Since the value of goods produced for own consumptiony plays an important role in Romania and could not be substracted from the consumption expenditure data, this value (EU-SILC variable HY170) has been added to the disposable income of Romanian households. Unfortunately, HY170 was not provided for 2020.

In order to reflect differences in household size and composition, the total disposable household income is "equivalised". An ‘equivalised disposable income’ for each member of the household is used in tables icw_pov_01, icw_pov_10, icw_pov_11 and icw_pov_12. It is based on a standard equivalence scale, the so-called “modified OECD” scale, which attributes a weight of 1.0 to the first adult in the household, 0.5 to each subsequent member of the household aged 14 and over and 0.3 to household members aged less than 14.

Consumption

Consumption expenditure data are taken from the Household Budget Survey (HBS) which describes monetary and non-monetary expenditures according to the Classification of individual consumption by purpose (COICOP) for each household. Total consumption expenditure is obtained by adding up the monetary (A) and non-monetary (B) expenditures of households for all COICOP items. For ICW purposes, "consumption" refers to monetary consumption expenditure. This means,

consumption = total consumption expenditure (CP00) minus non-monetary values [imputed rent (variable CP042) and other non-monetary expenditures (CP00B)].

An exception are 2010 data, for which 'consumption' includes non-monetary expenditures, but excludes imputed rent.

An ‘equivalised consumption’ is used for the ‘persons with low levels of expenditure’ (icw_pov_01) and ‘persons at two-fold risk of poverty’ (tables icw_pov_10, icw_pov_11 and icw_pov_12) indicators. As for income, equivalised consumption is obtained by dividing the total consumption of a household by the number of “equivalent adults” using the modified OECD scale. The resulting equivalised consumption is attributed to each member of the household.

Wealth

Households save and accumulate wealth as a precaution enabling them to smooth out a potential loss of income. Wealth data are collected by National Central Banks through the Household Finance and Consumption Survey (HFCS). The net wealth (variable DN3001) used in ICW statistics is derived from total household assets excluding public and occupational pension wealth (variable DA3001) minus total outstanding household liabilities (variable DL1000).  ‘Liquid financial wealth’ has been calculated as the sum of variables DA2101 value of deposits, DA2102 mutual funds, DA2103 bonds, DA2105 publicly traded shares and DA2106 managed accounts.

Share of households

The share of households, such as in table icw_res_01 "HH - Household", refers to the relative number of households in a sub-section of the population (for example households in the first quintile of the income distribution and at the same time in the first quintiles of the consumption expenditure and wealth distributions) measured as a percentage of all households in the population. Attention has to be paid to the fact that ICW statistics are built onto sample surveys: Only the households in the sample are divided into sub-sections, not the real population of households in the country. Thus, extrapolating the sample to the total population by multiplying each household with its sample weight, might change the share of households in the sub-section. This explains why the share of households in, for example, the first quintile of consumption (with the income and wealth quantile set to TOTAL) does not necessarily correspond to the exact 20% that would be expected.

Aggregate propensity to consume

Aggregate propensity to consume of the households is defined as follows:

where numerator is mean consumption and denominator is mean disposable income.

Persons at two-fold risk of poverty

Persons at two-fold risk of poverty are persons who are considered at risk of poverty in at least two of the following three dimensions: Income, consumption or liquid financial wealth.

Income poverty

The ‘at risk of income poverty rate’ is the share of people with an equivalised disposable income (after social transfer) below the ‘at risk of poverty threshold’, which is set at 60 % of the national median equivalised disposable income after social transfers:

where y refers to the household disposable income, e is the equivalisation factor of the household and  denotes the median for equivalised income at the individual level

Consumption poverty ( or persons with low levels of expenditures)

The definition of the at-risk-of-poverty indicator can be applied to consumption by considering individuals with low levels of expenditures those individuals i whose equivalised expenditures (c_i/e_i) are less than a threshold defined as follows:

where c refers to the total household consumption, e is the equivalisation factor of the household and  denotes the median for equivalised expenditures at the individual level.

In other words, a person is considered at risk of consumption poverty if his/her equivalised expenditures are below 60% of the median equivalised expenditure of all people in the country. ‘Low levels of expenditure’ and ‘consumption poverty’ are synonyms here.

Liquid financial wealth poverty

A threshold of below 60% of the median equivalised liquid financial wealth of all household members in the country defines the ‘at risk of liquid financial wealth poverty’:

where F refers to the total liquid financial wealth accumulated by the household, e is the equivalisation factor of the household and  denotes the median for equivalised liquid financial wealth at the individual level.

Household at risk of asset-based vulnerability

This indicator of vulnerability differs from the above in that it relates the equivalised amount of net wealth held by the household with the at risk of income poverty threshold. It is defined as follows:

where (w_i/e_i) is the equivalised amount of net wealth held by the household in which the individual i lives. µ denotes the scale of time (as a fraction of one year) and  stands for the usual median equivalised income.

The question is the following: for a given period, can the household maintain its material living conditions at a minimal level consuming its net wealth?

There is also the question of the nature of w_i. The debts and liabilities have of course to be included; the household can only spend his net wealth. In addition, the issue of liquidity is essential, as in a one-month scale for instance, the household cannot subsist if he only owns real-estate properties. The choice made here is not accounting for liquidity issues, but using net wealth as a whole. Thus, the concept may not reflect properly the concrete ability of the household to consume its assets swiftly if needed.

On the other hand, based on net wealth, the indicator addresses home-ownership, which plays an important role in poverty analysis. Indeed, home ownership entails an additional and implicit source of income, since home-owners benefit from the direct use of their houses; such additional source is not taken into account in the usual definition of disposable income at the micro-level.

Saving rate

Saving rates reflect the ability of households to spare money after having fulfilled their primary needs, regardless of the very nature of such needs. That means not presuming a given minimum level of expenses for a given household type with a given level of income.

The saving rate s_i for household i is defined as follows:

where y_i and c_i are respectively disposable income (variable HY020 in EU-SILC) and consumption (total consumption EUR_HE00 in HBS minus imputed rents EUR_HE042 minus other non-monetray expenditures EUR_HE00B).

Saving rates at the household level allow describing the distribution of saving flows among households, to compute the number of dissaving households (which could be seen as one indicator of vulnerability) and the mean and median saving rate. It is also possible to compute aggregate saving rates for multiple households as follows:

Dissaving household

It is a household whose consumption is higher than its income.

Essential services

Article 20 of the European Pillar of Social Rights demands access for everyone to essential services of good quality, including water, sanitation, energy, transport, financial services and digital communications. In our ‘affordability of essential services’ tables ‘icw_aff’, we define as essential services (SERV_ESS) the COICOP classes

• CP044 - CP0444 Water supply and miscellaneous services except other services relating to the dwelling n.e.c., which include CP0441 Water supply, CP0442 Refuse collection and CP0443 Sewage collection,
• CP045    Electricity, gas and other fuels
• CP073 - CP0733 - CP0734 Transport excluding air, sea and inland waterway, and
• CP083    Telephone and telefax services

We exclude financial services due to the low availability and quality of data.

Households and household members.

Individual private households. Institutional households and persons living in collective households or in institutions are generally excluded from the target population.

Indicators on income and consumption are available for different EU member states as follows:

2020: Belgium, Bulgaria, Czechia, Denmark, Germany, Estonia, Greece, Spain, France*, Croatia, Cyprus*, Latvia, Lithuania, Luxembourg, Hungary, Malta*, Netherlands, Austria, Poland, Romania, Slovenia and Slovakia. - Finland, Ireland and Portugal will follow. Cyprus, France, Malta and Sweden did not collect HBS data for the 2020 wave. Data for Italy 2020 were excluded due to missing income data in the original HBS data set, which make the statistical matching process unreliable.

* HBS data for Cyprus, France and Malta for the year 2020 have been extrapolated from the 2015 wave using the corresponding 2020 HICP coefficients.

2015: All EU member states except Italy (2015). Data for Italy 2015 were excluded due to missing income data in the original HBS data set, which make the statistical matching process unreliable.

2010: All EU member states except the Netherlands.  HBS 2010 data are missing entirely for the Netherlands.

Indicators on net wealth and liquid financial wealth are published for those countries covered by the HFCS (see section 3.8 on time coverage), except for Italy for the reason explained previously. It is important to note that HFCS income reference year for the 2015 estimations in Estonia and Portugal is 2012, while the EU-SILC income reference year is 2015 (see section 5 on reference period). This time lag requires estimates to be taken with caution.

The ICW experimental statistics cover three data collection waves: ‘around 2010’ , ‘around 2015’ and 'around 2020'. EU-SILC data corresponding to the same reference year around 2010, 2015 and 2020 as available in HBS are used for each country, and HFCS data with the closest wealth reference year available to that.

Not applicable.

Most indicators are reported as percentages (percentage of disposable income, percentage of total consumption, percentage of net wealth, percentage of gross income, percentage of total population, etc.). Table icw_res_02 is reported in Euro and Purchasing Power Standard (PPS). Table icw_sr_05 presents the Gini coefficient. Table icw_sr_06 is reported in thousand euro (THS_EUR) and PPS. If not specified elsewise, median values are reported.

The income reference period in EU-SILC is a fixed 12-month period (such as the previous calendar or tax year) for all countries except IE for which the survey is continuous and income is collected for the last twelve months. Other data is typically collected on the date of the survey.

HBS data are most often collected using expenditure diaries that respondents are asked to keep over a certain period of time (usually 1 or 2 weeks). These data are than extrapolated to refer to a full calendar year. The exact reference year of waves 2010, 2015 and 2020 for each country are displayed in the table below.

* HBS data for Cyprus, France and Malta for the year 2020 have been extrapolated from the 2015 wave using the corresponding 2020 HICP coefficients.

In the HFCS, the wealth reference year most often corresponds to the survey year (see table above). However, the income variables collected in this survey often refer to the year before. The following table shows the HFCS income reference year by country for the first and second wave:

 HFCS Income Reference Year 

Not applicable.

Not applicable.

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

Experimental ICW statistics are aggregate statistics that do not contain any confidential information. In addition, the microdata used in the ICW statistical matching are anonymized and do not contain any administrative information such as names or addresses that would allow direct identification. See also section 10.4. Dissemination format - microdata access.

All ICW statistics are Eurostat estimates and thus flagged ‘s'.

The EU-SILC publication rules were applied:

• Estimates have not been published if based on fewer than 20 sample observations.
• Estimate have been published with a flag ‘u – unreliable’ if based on 20 to 49 sample observations.
• Estimate have been published in the normal way when based on 50 or more sample observations.

There is no release calendar.

Not applicable.

In line with the Community legal framework and the European Statistics Code of Practice Eurostat disseminates European statistics on Eurostat's website (see item 10 - 'Accessibility and clarity') respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

The dissemination of ICW experimental statistics depends on the frequency of dissemination of the three sources needed for their computation:

• Income: EU-SILC has an annual frequency, with the income reference year being the year previous to the survey year. From 2020, EU-SILC includes the Overindebtedness, Comsumption and Wealth module every 3 years.
• Consumption: The frequency of HBS is every 5 years approximately: 1988, 1994 1999, 2005, 2010, 2015 and 2020. The next wave will be 2026.
• Wealth: The frequency of HFCS is every 3 years approximately: 2010, 2014, 2017 and 2021.

Due to the 5-yearly frequency of the HBS, joint ICW distributions can be produced only every 5 years.

No regular news release.

Diverse publications, such as Statistics Explained and methodological notes, are available on-line. There is a dedicated page at "Experimental Statistics".

Data are accessible online through Eurostat’s dissemination database under "Income, consumption and wealth - experimental statistics (icw)”.

Due to the confidential character of microdata and for not being the owner of underlying source microdata, no direct access to ICW microdata can be granted. Access to EU-SILC and HBS microdata can only be granted by the responsible entities following Eurostat’s access to microdata procedure. Access to HFCS microdata can only be granted by the ECB on behalf of the Household Finance and Consumption Network (HFCN). Access is in principle restricted to universities, research institutes, national statistical institutes and central banks inside the EU. Individuals cannot be granted direct access.

Contact point: estat-microdataaccess@ec.europa.eu

For more information refer to access to microdata

Internet address: http://ec.europa.eu/eurostat

For a detailed description of methods and concepts used, as well as for other documents related to the ICW experimental statistics, please consult the corresponding dedicated website.

See section 11.1 on quality assurance.

The ICW experimental statistics are not covered by a regulation and are computed by Eurostat based on microdata from EU-SILC, HBS and HFCS according to an agreed methodology. Please refer to the original sources (see links below in sections 'related metadata' and 'annexes') concerning the quality assurance of the underlying microdata.

 Several measures to assess the quality of the statistical matching are implemented, focusing for example on the

• Selection of the most suitable matching variables, assessing the comparability and similarity of the matching variables and their power to predict the variables of interest,
• Similarities between the original and the synthetic datasets regarding their marginal distributions of the variables of interest and the common variables,
• Similarities in the correlations observed in both datasets,
• Verification of the model’s assumptions.

Further details have been published in

• Lamarche, P., Oehler, F. and Riobóo, I. (2020), ‘European household’s income, consumption and wealth’, Statistical Journal of the IAOS, Vol.36/4, pp.1175-1188, https://doi.org/10.3233/SJI-190528
• Balestra, C. and Oehler, F. (2023), 'Measuring the joint distribution of income, consumption and wealth at micro level - methodological issues and experimental results', Eurostat Statistical Working Papers, 25 January 2023, https://ec.europa.eu/eurostat/web/products-statistical-working-papers/w/ks-tc-22-003

As for the original micro data, Eurostat does run a validation procedure of the microdata transmitted by National Statistical Institutes (EU-SILC and HBS) to check for any inconsistencies or missing data such that a minimum output quality standard is ensured. Likewise, the ECB is validating HFCS data transmissions. Please refer to those sources (see links below in sections 'related metadata' and 'annexes') concerning the quality reports that accompany the data to analyses their accuracy, coherence and comparability.

The joint distribution of income, consumption and wealth data establishes links between these three economic dimensions that define household’s material well-being and are at the focus of many data users.

The main users of the ICW experimental statistics are the following:

• Institutional users like other Commission services, other European institutions (such as the ECB), national administrations, or other international organisations like the Organisation of Economic Co-operation and Development (OECD).
• Statistical users at Eurostat or at EU National Statistical Institutes.
• Researchers.
• End users including policy makers interested in social indicators.

ICW statistics are new experimental statistics, so there is no information on user satisfaction yet.

The population covered by the ICW statistics is presented above under section 3.6, while the geographical coverage by year is explained in section 3.7.

The statistical matching reproduces the original distribution of total monetary consumption expenditure (HBS) very well in the matched dataset. The reproduction of the original distributin of net wealth (HFCS) is less accurate. Experimental indicators based on net wealth should be interpreted with care.

Moreover, it should be kept in mind that even if the statistical matching results into a good reproduction of the original distributions of income and consumption and wealth (of the source datasets) within the joint dataset, we cannot assess the accuracy of this joint distribution.

For accuracy concerning the original sources (EU-SILC, HBS and HFCS), please refer to their corresponding metadata (see links below in sections 'related metadata' and 'annexes').

Please refer to the original sources (see links below in section 'related metadata' and 'annexes').

Please refer to the original sources (see links below in sections 'related metadata' and 'annexes').

Once micro data for the corresponding reference year are available from EU-SILC, HBS and HFCS, the production of ICW statistics takes about 4 months.

EU-SILC micro data used to be available 11 months after the end of the survey year, which means 23 months after the end of the reference year. The timeliness is steadily improving and micro data are expected to be available 18 months after the end of the reference year in the future.

HBS micro data for most countries were available about 20 months after the end of the wave year. Some countries collected HBS data later and data were thus released several months later. Data are expected to be timelier from the 2026 wave.

HFCS micro data for the first (2010) wave were released in April 2013, for the second (2014) wave in December 2016, for the third (2017) wave in March 2020 and for the fourth (2021) wave in July 2023.

Not applicable.

The methodology applied to calculate ICW statistics is the same across all countries. However, it is important to note that the 2010, 2015 and 2020 data collection waves have in reality been carried out with different reference years by different countries. As such, the 2020 wave includes data for the reference year 2020 for Belgium Denmark, Estonia, Greece, Spain, Luxembourg, Hungary, Netherlands, Poland, Romania and Slovakia. The real reference year for Germany and Slovenia is 2018. Data for Bulgaria, Czechia, Latvia and Austria are mainly from 2019. For Austria, the HBS data collection period went from June 2019 to June 2020, for Bulgaria from April 2019 to March 2020. In Lithuania HBS data were collected from May 2021 to April 2022. Finland and Portugal were collecting HBS data in 2022 and Ireland from July 2022 to June 2023. For Cyprus, France and Malta HBS data for the year 2020 have been extrapolated from the 2015 wave using the corresponding 2020 HICP coefficients.

Further to that, the quality of EU-SILC, HBS and HFCS input data may differ among countries, and likewise the quality of the statistical matching of the three target values disposable income, (monetary) consumption expenditure and net wealth.

An exception in the concept of income and consumption has been made for Romania 2015: Since non-monetary consumption expenditure plays an important role in Romania, in particular for households with low income, but details on non-monetary expenditures were not provided and could therefore not be subtracted from the total consumption expenditure, we added the value of goods produced for own consumption to the disposable income. This is to correct at least partly for non-monetary expenditures contained for example in the saving rates for Romania. Due to a lack of the value of goods produced for own consumption in 2020, the correction could only be made for 2015. For all other countries, non-monetary expenditure is excluded from the total consumption expenditure and the value of goods produced for own consumption is excluded from income as stated under 3.4 statistical concepts and definitions.

It should also be kept in mind that different sets of variables may have been used to provide the link between income and consumption. These sets of "matching variables" are composed out of the following variables: Household size or household type (mandatory), degree of urbanisation, country of origin of the reference person, age of the refrence person, level of education of the reference person (detailed or aggregated), activity status of the reference person (detailed or aggregated) activity status of the reference person, occupation status of the reference person, tenure status or tenure/level of rent (mandatory), main source of income (detailed or aggregated), income ventiles (mandatory).

Wealth data have been added to the income-consumption micro dataset based on household size or household type, tenure status or tenure/level of rent, food consumption quintile and for all countries.

The methodology applied to calculate the ICW statistics ensures a comparison of reference years as far as source micro data are comparable over time. It is also important to note that the 2010, 2015 and 2020 data collection waves have in reality been carried out with different reference years by different countries (see section 15.1).

ICW experimental statistics are estimates computed by Eurostat. They follow international standards: ECOICOP, ISCO, ISCED, degree of urbanization and Canberra recommendations used in other social statistics domains. Best possible coherence with the respective publication of EU-SILC and HBS statistics has been sought. However, the one-dimensional indicators 'structure of the household population', 'gini', 'at risk-of-income poverty', 'mean and median income/consumption/net wealth by income/consumption/net wealth quantile' may differ from the original values published in the respective source data tables. This is due to methodological differences, such as for example the consumption concept, the definition of the household reference person, and the re-calibration of sample weights. Weights have been re-calibrated after the matching of income and consumption data to ensure a basic coherence of estimates with a set of socio-economic statistics published under both the ilc and hbs data tables (see ICW methodology). It is thus not recommended to use any one-dimensional ICW indicators, but to use ICW statistics uniquely for the added value of two-dimensional and three-dimensional indicators.

The calculation of EU aggregates follows the common practice of hbs and ilc tables (see 18.5).

Internal coherence among the original sources (EU-SILC, HBS and HFCS) used to compute ICW statistics is essential. As such, the coherence of the target and matching variables has been intensively analysed and is described in “ICW methodology” . Nevertheless, the coherence between EU-SILC disposable income (HY020) and consumption expenditure in HBS raises questions in a few specific cases like France and Greece 2015.

ICW statistics follow the guidelines for publication of EU-SILC and HBS statistics:

• An estimate should not be published if it is based on fewer than 20 sample observations
• An estimate should be published with a flag if it is based on 20 to 49 sample observations
• An estimate shall be published in the normal way when based on 50 or more sample observations

An exception is made for table icw_res_02 where estimates based on less than 30 sample observations are published with a flag 'u' for uncertainty. Table icw_res_01 shows the low share of households for some of the estimates in icw_res_02. Nevertheless, it was deemed important to allow users of these experimental statistics to see the full cross-over of income, consumption and wealth quantiles.

Not available.

The general Eurostat revision policy applies to this domain.

All reported errors (once validated) result in corrections of the disseminated data.

Reported errors are corrected in the disseminated data as soon as the correct data have been validated.

Data may be published even if they are missing for certain countries or flagged as provisional or of low reliability for certain countries. They are replaced with final data once transmitted and validated. European aggregates are updated for consistency with new country data. 

ICW statistics are derived from the statistical matching of three sources: the EU Statistics on Income and Living Conditions (EU-SILC), the Household Budget Survey (HBS) and the Household Finance and Consumption Survey (HFCS).

Please refer to the original sources (see links below in sections 'related metadata' and 'annexes') for additional information.

The frequency of data collection of the three sources needed for computing the ICW statistics is the following:

• EU-SILC: Annually
• HBS: Eurostat has collected and published individual country and European aggregate data every five years. However, about one third of the countries carry out annual surveys, another third have five-year intervals between surveys, and for the rest of the countries the periodicity varies.
• HFCS: There have been four survey waves, for which the data have been released in April 2013, December 2016, March 2020 and July 2023 respectively.

Please refer to the original sources (see links below in sections 'related metadata' and 'annexes').

Please refer to the original sources (see links below in sections 'related metadata' and 'annexes').

EU aggregates (EU and EU27_2020) are calculated as the population-weighted arithmetic average of individual country figures, except when shares of households are displayed. For shares of households, the actual share within the EU is calculated. EU aggregates are  displayed only if at least 70% of the total population is covered.

Please refer to the original sources (see links below in sections 'related metadata' and 'annexes').

No notes.