Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
The European Union Statistics on Income and Living Conditions (EU-SILC) is a survey-based instrument aiming at collecting timely and comparable cross-sectional and longitudinal multidimensional microdata on income, poverty, social exclusion and living conditions. In addition, it collects module variables every three years, six years or ad-hoc new policy needs modules.
The EU-SILC instrument provides two types of data:
Cross-sectional data pertaining to a given time or a certain time period with variables on income, poverty, social exclusion and other living conditions;
Longitudinal data pertaining to individual-level changes over time, observed periodically over four‐or more year rotation scheme (Annex III (2) of 2019/1700). Social exclusion and housing condition information is collected mainly at household level while labour, education and health information is obtained for persons aged 16 and over. The core of the instrument is income information at very detailed component level and mainly collected at personal level.
3.2. Classification system
International Standard Classification of Education (ISCED'2011);
International Standard Classification of Occupations (ISCO-08);
Classification of Economic Activities (NACE Rev.2-2008);
Common classification of territorial units for statistics (NUTS 2);
SCL - Geographical code list;
The recommendations made by the United Nations in the Canberra Group Handbook on Household Income Statistics should also be taken into account.
For more details on the classification used please, see RAMON, Eurostat's metadata server.
3.3. Coverage - sector
Data refer to all private households and individuals living in the private households in the national territory at the time of data collection.
The EU-SILC survey is a key instrument for the European Semester and the European Pillar of Social Rights, providing information on income distribution, poverty and social exclusion, as well as various related living conditions and poverty EU policies, such as on child poverty, access to health care and other services, housing, over indebtedness and quality of life. It is also the main source of data for microsimulation purposes and flash estimates of income distribution and poverty rates.
Further details are provided in items 5, 15.1.1.1, 15.2.2 and 18.3.
The source or procedure used for the collection of income variables
The Danish data on incomes are mainly collected in gross form from registers, stemming from the Danish Tax Authoritie. The only exception is intra-household transfers, that are collected via survey.
Many variables are therefore partly based on Danish tax regulation by design. For this reason, a number of adjustments are made to the variables in order to align the data to the SILC definition of incomes. This is possible for most variables, see a more thorough description on the publication 'revision of SILC incomes'.
The tax paid is subtracted from the gross household income to produce the net household income. Because the Danish tax system is complex and almost all income and deductions are part of a integrated tax calculation, it is not possible to calculate person tax rates for all net income components. The total household income is then distributed among the net components the following way. All income components that are exempt from tax are subtracted from the gross income. Tax on pensions from indicidual private plans are taxed separately, this tax is removed from both the gross income and the tax paid, to be added later. The adjusted tax paid and gross income are then used to calculate an effective tax rate at the person level. The tax rate is fixed between 0 and 100 percent of the gross income, any tax outside of this range is attributed to income components according to where the largest gross incomes are found. Typically negative tax rates and tax rates above 100 are found for self-employed persons, so most of this excess tax is placed in PY050N.
Comparability and deviation from definition for each income component
All income components are fully comparable
Household definition and membership
Reference population - Persons living in private households on Danish territory at the end of the income reference period Private household - A person living alone or a group of persons who live together, providing oneself or themselves with the essentials of living Household membership - A person residing in the household or being temporarily away from the household and sharing the cost of living to a large degree
Description of reference period used for incomes
Period for taxes on income and social insurance contributions: 01 January 2023 - 31 December 2023 Income reference periods used: 2023 Reference period for taxes on wealth: 31 December 2023 Lag between the income ref period and current variables: 2-5 months, depending on the date of the interview. Register data for March are typically used for current register variables
3.5. Statistical unit
Statistical units are private households and all persons living in these households who have usual residence in the Member State. Annex II of the Commission implementing regulation (EU) 2019/2242 defines specific statistical units per variable and specifies the, content of the quality reports on the organization of a sample survey in the income and living conditions domain pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council.
There are two types of general statistical units in the SILC survey.
Persons - A person living in a private household at the end of the income reference period or a current household member of the sampled person
Households - The entire household of the sampled person
Specific statistical units per variable are defined in Annex II of the Commission implementing regulation (EU) 2019/2242 specifying content of the quality reports on the organization of a sample survey in the income and living conditions domain pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council.
3.6. Statistical population
The target population is private households and all persons composing these households having their usual residence in the Member State. Private household means a person living alone or a group of persons who live together, providing oneself or themselves with the essentials of living.
3.6.1. Reference population
Definitions of reference population, household and household membership
Reference population
Private household definition
Household membership
Persons with their legal address within Denmark
Persons living in private households and thereby excluding persons living in institutions, prison or the homeless.
The composition of the household with regards to household membership is defined by the respondent at the time of interview and is based upon information about who lives on the address and whether or not they share expenses on living conditions and food
3.6.2. Population not covered by the data collection
The sub-populations that are not covered by the data collection includes: those who moved out of the country’s territory; or those with no usual residence; or those living in institutions or who have moved to an institution compared to the previous year.
3.7. Reference area
Denmark
3.8. Coverage - Time
Reference periods
Income variables: 2023.
Current variables: 2024.
The data are collected annualy. The interviews were conducted in the period 06 March 2024 - 26 May 2024.
Data are available for the survey years 2004-2024.
3.9. Base period
Not applicable.
The data involves several units of measure depending upon the variables. Income variables are transmitted to Eurostat in national currency. For more information, see methodological guidelines and description of EU-SILC target variables available on CIRCABC
Description of reference period used for incomes
Period for taxes on income and social insurance contributions
Income reference periods used
Reference period for taxes on wealth
Lag between the income ref period and current variables
2023
2023
2023
3-5 months lag depending on interview date
6.1. Institutional Mandate - legal acts and other agreements
Regulation (EU) 2019/1700 was publish in OJ on 10 October 2019, establishing a common framework for European statistics relating to persons and households, based on data at individual level collected from samples (IESS). The Annex to the Commission implementing regulation (EU) 2019/2180 of 16 December 2019 specifies the detailed arrangements and content for the quality reports pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council and Regulation (EU) 2019/2242.
6.2. Institutional Mandate - data sharing
Confidential microdata are not disclosed by Eurostat. Access to confidential microdata for scientific purposes may be granted on the basis of Commission Regulation 557/2013 and Regulation 223/2009 of the European Parliament and the Council on European statistics.
7.1. Confidentiality - policy
All personal identifiers have been removed from the data. De-identified microdata can be acquired through Eurostat and Danish research institutions can apply for access to Danish data through Statistics Denmark research services.
Only aggregated data and tables are disseminated in publications and on the web. Standard discretionary measures are taken for the dissemination of data.
7.2. Confidentiality - data treatment
All microdata are kept confidential and use thereof is restricted to the few persons producing the SILC. All identifying information (names and CPR-numbers) are removed as soon as possible and psueodonomized id-numbers are used in stead.
EU-SILC is primarily survey based and the results are therefore subject to statistical uncertainty. The uncertainty of the cross-sectional indicators on the full population is quite low. However, the statistical uncertainty becomes significant, when looking at small sub-populations due to the relatively small sample size and high variation in the sample weights.
Statistic Denmark recommends the use of national register data for users who are only looking for statistics concerning Denmark and is not looking to do cross-country comparisons. Statistics Denmark registers cover most of the objective SILC variables on demographics, housing, labour market participation, education, income and wealth. For access to register data, please find more information on the Statistics Denmark website.
12.1. Relevance - User Needs
The main users of EU-SILC statistical data are policy makers, research institutes, media, and students.
12.2. Relevance - User Satisfaction
Eurostat carried out an online general User Satisfaction Survey (USS) in the period between April and July 2019 to obtain a better knowledge about users, considering their needs and satisfaction with the services provided by Eurostat. The survey has shown that EU-SILC is of very high relevance for users. For the majority, both aggregates and micro-data were important or essential in their work irrespective of the purpose of their use. The use of the ad-hoc modules was less widespread than the use of the nucleus variables. Nevertheless, there was high interest to repeat these modules in order to have the possibility of comparing data over time. Users emphasized their strong need for more detailed micro-data, which is currently not possible. Under the new legal framework implemented from 2021, the NUTS 2 division will be available for the main indicators. Finally, users were satisfied with overall quality of the service delivered by Eurostat, which encompasses data quality and the supporting service provided to them.
Data for respondents in the SILC-survey is generally complete and in accordance with the SILC guidelines. As the more complex data on incomes, education and housing stems from full population registers, the item non-response is very low in Denmark. For most questions in the webquestionnaire item non-response is hidden by default and only shown if the respondent tries to go to the next question without answering the current question.
The following variables from the nucleus are not collected: HY170G HY145.
With regards to HY145: by the time we receive the income data from our income register tax paybacks have already been calculated and are therefor equal to 0.
The following variables are optional and are not collected: HY030G, RL080.
DB060, DB062, DB070 and HS022 are not relevant in a Danish context.
Longitudinal weights for 5 and 6 years and household grid variables related to more than 12 members of the household are not relevant in the Danish SILC setup.
The following variables differ from the definition:
PY030G: Historically there has been a challenge for Denmark to distinguish voluntary from statutory pension contributions. How to best solve this is being looked into bilaterally between Denmark and Eurostat but for the time being the variable remains empty.
12.3.1. Data completeness - rate
Not required
13.1. Accuracy - overall
According to Reg. (EU) 2019/1700 Annex II, precision requirements for all data sets are expressed in standard errors and are defined as continuous functions of the actual estimates and of the size of the statistical population in a country or in a NUTS 2 region. For the income and living conditions domain, the estimated standard errors of the following indicators are examined according to certain parameters set:
· Ratio at‐risk‐of‐poverty or social exclusion to population;
· Ratio of at‐persistent‐risk‐of‐poverty over four years to population;
· Ratio at‐risk‐of‐poverty or social exclusion to population in each NUTS 2 region.
Further information is provided in section 13.2 Sampling error.
Possible sources of bias
Population definition - Some bias may arise when we remove households with 10 adults or more.
Contact at household - Some households cannot be located. If those households differ from the households that can be contacted, a bias will arise
Non-response bias - The persons who refuse to participate in the survey are not representative of the sample as a whole. Typically, younger individuals and individuals with lower incomes are more inclined to refuse to participate in the SILC survey.
These problems are partly mitigated through the calibration of weights (see section 18.5). The calibration ensures that AROP60 almost exactly matches the register, so variables that are closely related to the AROP60 have a lower selection bias than variables that are not that related to AROP60.
Steps are continuously taken in order to reduce the non-response rate, which is the largest source of bias effecting the results.
13.2. Sampling error
EU-SILC is a complex survey involving different sampling designs in different countries. In order to harmonize and make sampling errors comparable among countries, Eurostat (with the substantial methodological support of Net-SILC2) has chosen to apply the "linearization" technique coupled with the “ultimate cluster” approach for variance estimation.
Linearization is a technique based on the use of linear approximation to reduce non-linear statistics to a linear form, justified by asymptotic properties of the estimator. This technique can encompass a wide variety of indicators, including EU-SILC indicators. The "ultimate cluster" approach is a simplification consisting in calculating the variance taking into account only variation among Primary Sampling Unit (PSU) totals. This method requires first stage sampling fractions to be small which is nearly always the case. This method allows a great flexibility and simplifies the calculations of variances. It can also be generalized to calculate variance of the differences of one year to another.
The main hypothesis on which the calculations are based is that the "at risk of poverty" threshold is fixed. According to the characteristics and availability of data for different countries, we have used different variables to specify strata and cluster information.
In particular, countries have been split into 3 groups:
BE, BG, CZ, IE, EL, ES, FR, HR, IT, LV, HU, PL, PT, RO, SI, UK and AL, whose sampling design could be assimilated to a two-stage stratified type we used DB050 (primary strata) for strata specification and DB060 (Primary Sampling Unit) for cluster specification;
DK, DE, EE, CY, LT, LU, NL, AT, SK, FI, CH whose sampling design could be assimilated to a one stage stratified type we used DB050 for strata specification and DB030 (household ID) for cluster specification;
MT, SE, IS, NO, whose sampling design could be assimilated to a simple random sampling, we used DB030 for cluster specification and no strata.
13.2.1. Sampling error - indicators
The concept of accuracy refers to the precision of estimates computed from a sample rather than from the entire population. Accuracy depends on sample size, sampling design effects and structure of the population under study. In addition to that, sampling errors and non-sampling errors need to be taken into account. Sampling error refers to the variability that occurs at random because of the use of a sample rather than a census and non-sampling errors are errors that occur in all phases of the data collection and production process.
AROPE
At risk of povery -60%
Severe material depreivation
Very low work intensity
Ind. Value
Stand. Errors
95 % CI Lower
95 % CI Upper
Ind. Value
Stand. Errors
95 % CI Lower
95 % CI Upper
Ind. Value
Stand. Errors
95 % CI Lower
95 % CI Upper
Ind. Value
Stand. Errors
95 % CI Lower
95 % CI Upper
Total
18,0
0,7
16,6
19,4
11,6
0,6
10,5
12,8
4,0
0,4
0,1
3,2
10,6
0,6
9,3
11,8
Male
17,8
0,8
16,1
19,4
11,6
0,7
10,2
13,0
3,9
0,4
3,1
4,8
10,3
0,8
8,8
11,7
Female
18,2
0,8
16,7
19,8
11,7
0,6
10,5
13,0
4,0
0,4
3,3
4,8
10,9
0,7
9,5
12,3
Age 0-17
15,9
1,7
12,7
19,2
10,1
1,4
7,3
12,9
4,7
0,9
2,8
6,6
6,2
1,1
4,1
8,3
Age 18-64
20,7
0,8
19,1
22,2
12,3
0,7
11,1
13,6
4,6
0,4
3,8
5,3
12,1
0,6
10,9
13,3
Age 65+
12,3
0,6
11,1
13,6
11,1
0,6
9,9
12,3
1,5
0,2
1,1
2,0
NA
NA
NA
NA
NUTS 2 (Region Hovedstaden)
17,1
1,2
14,7
14,7
11,3
1,0
9,4
13,3
4,0
0,8
2,5
5,5
11,6
1,2
9,2
14,0
NUTS 2 (Region Sjælland)
15,1
1,5
12,2
18,0
9,1
1,2
6,8
11,4
4,9
1,0
2,9
6,9
6,3
0,9
4,6
8,0
NUTS 2 (Regions Syddanmark)
18,5
1,6
15,4
21,5
11,5
1,4
8,8
14,1
3,8
0,7
2,3
5,3
10,0
1,1
7,9
12,0
NUTS 2 (Region Midtjylland)
19,7
1,7
16,4
23,0
13,0
1,4
10,2
15,7
3,5
0,7
2,1
4,9
11,7
1,7
8,4
15,0
NUTS 2 (Region Nordjylland)
20,3
2,0
16,4
24,1
13,8
1,8
10,2
17,3
3,9
0,8
2,4
5,5
12,0
1,5
9,0
15,1
Persistent risk of poverty 2024
Country
Percent
Std Err of
Variance of
95% Confidence Limits
Percent
Percent
for Percent
DK
**This table will be filled out once the longitudinal weights have been calculated
Coverage errors: errors due to divergences existing between the target population and the sampling frame.
Measurement errors: errors that occur at the time of data collection. There are a number of sources for these errors such as the survey instrument, the information system, the interviewer and the mode of collection.
Processing errors: errors in post-data-collection processes such as data entry, keying, editing and weighting.
Non-response errors: errors due to an unsuccessful attempt to obtain the desired information from an eligible unit. Two main types of non-response errors are considered:
Unit non-response: refers to absence of information of the whole units (households and/or persons) selected into the sample.
Item non-response: refers to the situation where a sample unit has been successfully enumerated, but not all required information has been obtained.
13.3.1. Coverage error
Coverage errors include over-coverage, under-coverage and misclassification:
Over-coverage: relates either to wrongly classified units that are in fact out of scope, or to units that do not exist in practice.
Under-coverage: refers to units not included in the sampling frame.
Misclassification: refers to incorrect classification of units that belong to the target population
Known coverage errors
Persons and households entering the country between the 31st of December and the start of the interview period in February are not part of the sampling frame.
If a person, who belongs to a household from a still active SILC panel, is selected, the new household is dropped. This situation is primarily of theoretical interest. The practical importance is negliaible.
Only private households are included in SILC. Statistics Denmark does not include any households consisting of 10 or more adults in the survey sample. The same households are not excluded from the final callibration.
13.3.1.1. Over-coverage - rate
Coverage error
Main problems
Population (sub-population)
Size of error
Comments
Over-coverage
Persons added to the population (from 1st of January to 31st of March)
0,8
Under-coverage
People registered in adresses with more than 10 adults
0,7
Misclassification
Not available
13.3.1.2. Common units - proportion
Not required
13.3.2. Measurement error
Measurement error for cross-sectional data
Cross-sectional data
Source of measurement errors
Building process of questionnaire
Interview training
Quality control
Some questions are subjective in nature.
Respondents might interpret questions differently from others or not be able to remember certain aspects.
Register data on income are not created to match the SILC manual perfectly. Thus there are a few deviations from the manual.
Two seperate questionnaires are formed. A Web Questionnaire and one for Telephone interviews. Apart from slight differences in the introductory texts they are mostly alike.
The web interviews include a suggestion box, where bad and good experiences are collected in order to improve next years' survey.
Telephone interviewers have a 2 hour instruction followed by test-interviews with each other and Q&A sessions
Quality control was ensured by Statistics Denmark Staff
13.3.3. Non response error
Non-response errors are errors due to an unsuccessful attempt to obtain the desired information from an eligible unit. Two main types of non-response errors are considered:
1) Unit non-response which refers to the absence of information of the whole units (households and/or persons) selected into the sample. According to Annex VI of the Reg.(EU) 2019/2242
Household non-response rates (NRh) is computed as follows:
NRh=(1-(Ra * Rh)) * 100
Where Ra is the address contact rate defined as:
Ra= Number of address/selected person (including phone, mail if applicable) successfully contacted/Number of valid addresses/selected person (including phone, mail if applicable) selected
and Rh is the proportion of complete household interviews accepted for the database
Rh=Number of household interviews completed and accepted for database/Number of eligible households at contacted addresses (including phone, mail if applicable)
• Individual non-response rates (NRp) is computed as follows:
NRp=(1-(Rp)) * 100
Where Rp is the proportion of complete personal interviews within the households accepted for the database
Rp= Number of personal interview completed/Number of eligible individuals in the households whose interviews were completed and accepted for the database
• Overall individual non-response rates (*NRp) is computed as follows:
*NRp=(1-(Ra * Rh * Rp)) * 100
For those Members States where a sample of persons rather than a sample of households (addresses, phones, mails etc.) was selected, the individual non-response rates will be calculated for ‘the selected respondent.
2) Item non-response which refers to the situation where a sample unit has been successfully enumerated, but not all the required information has been obtained.
13.3.3.1. Unit non-response - rate
Unit non-response rate for cross-sectional
Address (including phone, mail if applicable) contact rate
Complete household interviews
Complete personal interviews
Household Non-response rate
Individual non-response rate
Overall individual non-response rate
(Ra)
(Rh)
(Rp)
(NRh)
(NRp)
(NRp)*
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
99,69
99,48
100,00
48,89
34,64
73,78
100,00
100,00
100,00
51,27
65,54
26,22
0,00
0,00
0,00
51,27
65,54
26,22
where
A=total (cross-sectional) sample,
B =New sub-sample (new rotational group) introduced for first time in the survey this year,
C= Sub-sample (rotational group) surveyed for last time in the survey this year.
13.3.3.2. Item non-response - rate
The computation of item non-response is essential to fulfil the precision requirements. Item non-response rate is provided for the main income variables both at household and personal level.
Item non-response which refers to the situation where a sample unit has been successfully enumerated, but not all the required information has been obtained.
13.3.3.2.1. Item non-response rate by indicator
Due the the source of the income variables being the income register, there is no item non-response issues regarding these variables.
The following variables have an item non-response of more than 5 percent:
(c) Individuals not interviewed for reasons other than their being out of scope %
0,0 %
0,0 %
0,0 %
Not applicable
Not applicable
Fill in if more than 4 rotation panel is used in your country
Part II
Re-interview rates
Wave 2
Wave 3
Wave 4
Wave 5
Wave6
Re-interview rates for people leaving their household
Total
4,9 %
5,4 %
3,8 %
Not applicable
Not applicable
Males
5,0 %
4,8 %
4,0 %
Not applicable
Not applicable
Females
4,9 %
5,7 %
3,7 %
Not applicable
Not applicable
Re-interview rates for young people (16-35) leaving their household
Total
9,6 %
12,9 %
5,8 %
Not applicable
Not applicable
Males
10,2 %
11,4 %
5,4 %
Not applicable
Not applicable
Females
9,2 %
14,4 %
6,3 %
Not applicable
Not applicable
Fill in if more than 4 rotation panel is used in your country
a) and c) Calculated on household level
b)Calculated on individual level
13.3.5. Model assumption error
In the 2014 revision a cap on negative incomes at €100.000 has been introduced into the calibration, in order to increase comparability with other European countries. Any indivduals with negative income surpassing this amount is still included in the SILC, but will have lower weights.
Negative incomes that stems from losses for self-employed, windfall capital losses and interest expenses is fully covered in the danish SILC in the year the losses is endured. This might still lead to a slight overestimation of some indicators of inequality comparatively to countries, where incomes are bottom coded.
14.1. Timeliness
The micro data were submitted to Eurostat by the end of the survey year and less than 12 months after the end of the income reference period.
14.1.1. Time lag - first result
12 months after end of the income reference period - data was transmitted to Eurostat by the end of the year of data collection. National results based on cross-sectional data were released to coincide with this delivery in December of 2024. Final income data were transmitted with the first data transmission
14.1.2. Time lag - final result
See 14.1.1 - as the initial data collection did not contain any provisional income data but rather the final income data the results did not change
14.2. Punctuality
Data was delivered on time
14.2.1. Punctuality - delivery and publication
National data were released on December 18th 2024. On December 23rd data was transmitted to Eurostat some seven days prior to deadline.
15.1. Comparability - geographical
Data is generally comparable within Europe. But caution is advised, when interpreting minor differences on indicators between countries due to statististical errors and the difficulties related to cross-border income comparisons.
15.1.1. Asymmetry for mirror flow statistics - coefficient
Not applicable.
15.2. Comparability - over time
Denmark has participated in the EU-SILC since 2003.
In 2021, the SILC survey was aligned to the EU Regulation 2019/1700 (IESS Regulation - find more information legal framework). This has introduced some new variables and resulted in changes to some other variables. If there is a breaking change, the variable has received a new name. Therefore there are no known data breaks due to the IESS Regulation.
In 2020, there was a major revision in the method behind the calculation of income variables. Read the detailed description revision of SILC incomes.
In 2018, there was a error in the first days of the data collection. This resulted in 405 rejected households due to a looping issue that caused some household members not to be interviewed. 276 of these households was included in SILC. The labour markets status of the missing houshold members was imuted via income data from the preliminary income register and answers provided in previous years. The issue has been described in more detail in an annex to the quality report.
In 2017, parental leave payments was moved from PY120 to HY050
In 2013, a policy change on private lump-sum pensions, has lead to an increase in private pension pay-outs. The effect is temporary only. As SILC 2014 contains incomes from 2013, the SILC-2014 is the first year affected by this policy-change. The expected effect on gini is
SILC-2013: +0
SILC-2014: +0.31
SILC-2015: +0.25
SILC-2016 and forth: ~-0.1 (Provided that we are unable to impute the lump-sum pensions).
Threshold indicators such as Risk of Poverty should be virtually unaffected as most people with private lump-sum pensions have incomes above the median.
In 2014 and again in 2016, Statistics Denmark revised the calibration of SILC. The new calibration has been implemented going back to 2011.
From 2011, the income mass within income groups has been included in the calibration to fit the register better. This has been done in order to obtain better consistency between our register data and the EU-SILC data and has significantly lowered the deviations between full register data and the silc data, when measuring average income and the gini-coefficient. Furthermore between 2009 and 2010 the household definition in the callibration changed from adresses to a narrow concept of the family.
For consistent data on the gini-coefficient and similar pure economic indicators pre-2010, it's recommended to use statistics based on Danish register data for the entire population.
Time series are comparable since 2011, with the exception of incomes which exhibit a data break in 2020.
2003-2011 has some breaks in 2009, 2010 and 2011. These are mainly related to a change in the calibration, to ensure that the effcts of financial crisis was easured accurately and changes in the household definition used in the calibration.
15.2.2. Comparability and deviation from definition for each income variable
Comparability and deviation from definition for each income variable
Income
Identifier
Comparability
Deviation from definition if any
Total hh gross income
(HY010)
F
Total disposable hh income
(HY020)
F
Total disposable hh income before social transfers other than old-age and survivors' benefits
(HY022)
F
Total disposable hh income before all social transfers
(HY023)
F
Income from rental of property or land
(HY040)
L
Income from renting out property which is exempt for taxation (less than 32.500 DKK pr year) is currently not part of the data
Family/ Children related allowances
(HY050)
F
Social exclusion payments not elsewhere classified
(HY060)
F
Housing allowances
(HY070)
F
Regular inter-hh cash transfers received
(HY080)
F
Alimonies received
(HY081)
F
Interest, dividends, profit from capital investments in incorporated businesses
(HY090)
F
Interest paid on mortgage
(HY100)
F
Income received by people aged under 16
(HY110)
F
Regular taxes on wealth
(HY120)
F
Taxes paid on ownership of household main dwelling
(HY121)
F
Regular inter-hh transfers paid
(HY130)
F
Alimonies paid
(HY131)
F
Tax on income and social contributions
(HY140)
F
Repayments/receipts for tax adjustment
(HY145)
NC
By the time we receive the income data from our income register tax paybacks have already been calculated and are therefor equal to 0
Value of goods produced for own consumption
(HY170)
NC
This is not information that is available in the Danish registers and is deemed to be of very seldom and non-significant occurance that it is not included.
Cash or near-cash employee income
(PY010)
F
Other non-cash employee income
(PY020)
F
Income from private use of company car
(PY021)
F
Employers social insurance contributions
(PY030)
NC
Most social contributions from employers are not mandated by law - however contributions in accordance with voluntary labour market aggreements are common - but not included in PY030). ATP contributions will be included from 2025. How exactly to best solve this variable is being looked into bilaterally between Denmark and Eurostat but for 2024 the variable remains empty.
Contributions to individual private pension plans
(PY035)
F
Cash profits or losses from self-employment
(PY050)
F
Pension from individual private plans
(PY080)
F
Unemployment benefits
(PY090)
F
Old-age benefits
(PY100)
F
Survivors benefits
(PY110)
F
Sickness benefits
(PY120)
F
Disability benefits
(PY130)
F
Education-related allowances
(PY140)
F
F= Fully comparable; L= Largely comparable; P= Partly comparable and NC= Not collected.
15.3. Coherence - cross domain
The coherence of two or more statistical outputs refers to the degree to which the statistical processes, by which they were generated, used the same concepts and harmonised methods. A comparison with external sources for all income target variables and the number of persons who receive income from each ‘income component’ will be provided, where the Member States concerned consider such external data to be sufficiently reliable.
15.3.1. Coherence - sub annual and annual statistics
Not applicable.
15.3.2. Coherence - National Accounts
Information provided in Annex 7*
Annex 7 requires longitudinal weights calculated in order to be generated. Annex 7 will be uploaded once the longitudinal weights for the 2024 Danish SILC are ready.
15.4. Coherence - internal
There are slight differences between aggregated SILC and the National Danish register variables on incomes.
Imputed rent
Interests payments related to the mortgage
Fringe benefits (except value of company car)
Are not inclued in the disposable income in SILC, but are part of disposable income in National Danish Statistics.
Lump-sum private pension payouts are included in the SILC disposable income, but are not included in National Danish Statistics on incomes.
Mean (average) interview duration per household = 28.5 minutes.
Mean (average) interview duration per person = 8.6 minutes.
Mean (average) interview duration for selected respondents (if applicable) = 14.2 minutes.
17.1. Data revision - policy
The SILC results are communicated nationally at the end of the survey year. Following this, micro-data are transmitted to Eurostat. The data needs to pass validation when being transmitted. Once transmitted Eurostat review the data. The review can lead to minor changes being made in the data. When the data is accepted by Eurostat, it is marked as final.
17.2. Data revision - practice
National results were first disseminated in December 2024. There were no revisions
17.2.1. Data revision - average size
No revisions
Detailed information concerning sampling frame, sampling design, sampling units, sampling size, weightings and mode of data collection can be found in this section (please see below). Such information is mainly used for the computation of the accuracy measures.
18.1. Source data
Data on household composition, current labour market participation and subjective questions are based on interviews. Objective data on housing, education, basic demographics and incomes are based on administrative registers where available.
Survey data
The variables that either cannot or may not be collected from registers are collected via a survey. The survey is designed to meet the definitions laid out in the regulation governing each survey year, customized to fit a national context. The data collection is carried out in spring, by telephone and web.
Register data
A large part of the SILC survey relies on register data. In general, if a register exists that corresponds with a EU-SILC variable, the register is used instead of adding said variable to the survey. The majority of variables related to income, wealth, population and the dwelling are covered by register. These are the main registers used in the compilation of the SILC variables.
Population register (CPR) - Centrally administered population register containing information on the Danish population, with information on sex, age, id's of fathers, mothers and possible spouses, home addresses etc.
Income register - Register of incomes from the Danish tax authority covering all persons paying tax in Denmark
Wealth register - Register of wealth, collected from numerous sources. Covers most financial and non-financial assets and liabilities of Danish individuals and households
Education register - Data from administrative records of the edicational level and activities of the population
Building register - A register of residential and non-residential buildings in Denmark, containing details on the use of the building and the building itself
e-Income register - A register of incomes sourced from the Danish tax authority. Also contains hours worked and taxable benefits
Various government benefit registers - Numerous smaller registers containing information on social benefits have been used in the compilation
Some information from other registers have been used, but to a lesser degree.
18.1.1. Sampling Design
EU-SILC has a rotating panel structure, where the four panels constitute the cross-sectional sample. A panel is invited to participate four consecutive years, after which it is replaced by a fresh sample constituting a new first panel. From 2016, the Danish sample design has been a stratified one-stage clustered sample; the clusters being individual SILC-families.
From each cluster an individual is randomly selected: the contact person. Denmark uses the Selected Respondent model, where this individual answers the relevant questions for all household members.
A sample size (number of clusters) for the new panel is determined at the beginning of the sampling process. The sample size is then allocated to the strata, with respect to both the AROPE standard error and the number of households in the strata compared to the number of sampled households in each strata.
The population is stratified by income and NUTS-2 regions, in order to limit the standard errors in smaller regions and lower income groups. The income for the entire previous year is not known at the time of selection, which is why an estimate for the annual income is produced, based on income for the first eleven months of the year before, combined with information on non-wage income for the year prior to the year before. These estimates are then grouped into three strata; 0-60 pct. of the median income, 60-100 pct. of the median income and 100+ of the median income. These income brackets are then permutated with the five NUTS-2 regions to produce 15 unique strata. This stratification method was implemented in the SILC-2020 survey year. Before that, the strata was made purely on the NUTS-2 regions.
18.1.2. Sampling unit
SILC-family is the primary sampling unit (PSU), and the household members of the contact person constitute the cluster. In some respects the contact person is the secondary sampling unit.
18.1.3. Sampling frame
The sampling frame for the PSUs is constructed from the Central Personal register (CPR). This includes all persons in the population at the end of the income reference period and is at the time reduced to persons in private households. We do not have a register of persons in non-private households, so this is done by approximation. All addresses where 10 or more adults are residing are removed from the sample frame, as these households have a large probability of being institutions or other administrative households and therefore would fall outside of the scope of the statistical population.
An analysis from September 2021 showed, that there were 1072 addresses with 10+ residents, containing around 28,000 residents. Around 5,000 of these can be identified as homeless and around 1,500 are temporarily living abroad in other nordic countries. Both these groups are registered at a common (fictional) address and therefore show up as large households in the register. There is currently no reliable data that can identify the household type of the remaining residents, but they consist of institutions, prisons, student housing, housing for the elderly and communal living arrangements.
Contact persons are 16 years of age or older. The sampling of contact persons takes the panel aspect into account. Thus the sample includes persons turning 16 in the panel duration as well. They are however not survey until they turn 16. The sampling frame for the contact persons consisted of persons aged 13 years or older and registered in households/addresses from the PSU sampling frame.
Persons that refuse to participate in a wave, are not contacted in the following wave.
Sample size
Number of households in the sample: 12,428
Number of households in the sample, excl. dead, children under 16 years old and non-private households: 12,294
Actual sample size (contacted households): 12,428
Achieved sample size: 6,010
Population size
Number of persons in entire population: 5,961,249
Number of persons aged 16 or above in population: 4,952,833
Number of families (proxy for households) in population: 3,074,864
18.2. Frequency of data collection
Annual data collection. Data collection is conducted between February and May of the year following the reference year
18.3. Data collection
Mode of data collection
1-PAPI
2-CAPI
3-CATI
4-CAWI
5-PAPI proxy
6-CAPI-proxy
7-CATI-proxy
8-CAWI proxy
9-other
% of total
13,7
83,1
0,9
0,5
Description of collecting income variables
The source or procedure used for the collection of income variables
The form (gross, net) in which income variables at component level have been obtained
The method used for obtaining target variables in the required form
Register data (except for inter-household transfers)
Gross
Using income register data to produce income variables
During the interviews, filters ensure that only valid answers to the questions can be given. When entering amounts, the respondent will recieve a warning and given the possibility to change his/her answer if the answer is significantly outside the norms.
After the interview process a range of basic checks for consistency is carried out and data is adjusted when deemed neccesary. These include, but are not limited to, outlier detection, missing value analysis, distribution analysis and cross-validation of related variables.
Only households where Statistics Denmark is able to identify the personal ID (CPR) for all adult members is included in the final survey data.
18.5. Data compilation
Data compilation is carried out in several steps, as described below. The first steps of the process only pertain to cross-sectional data. Later, data from previous years are added and longitudinal weights are calculated.
Raw data manipulation and sample update
Data are received from our data collection partner via a secure FTP-server. Data is then transformed to be more compatible with SILC data formats and duplicates are removed. The sample is updated with metadata on the interview and all persons are assigned personal id's (PB030/RB030). The consolidated sample is then used to create the D, H, R and P populations.
Survey and register variables
The data from the survey is then extracted and recoded to comply with modalities in EU-SILC regulation. Seperate programs extract data from registers and edit to comply as well.
Collection of data, weighting and flag coding
When all variables are produced, they are collected into the D, H, R and P files. Flags are then coded according to regulation and added to the respective files. Weights are then calculated on both a personal and household level.
Longitudinal data
When the cross-sectional data is produced, the data from tre previous three years are gathered from last years transmission. This collected data is used to produce longitudinal weights and ultimately the final EU-SILC dataset.
18.5.1. Imputation - rate
Statistics Denmark does not impute values for any individual observations, mainly due to the fact that the income register is complete and there are no missing values that would hinder the correct calculation of total incomes.
18.5.2. Calculation of weighting factors and weight adjustments
Weights were constructed by means of the generalized regression estimator. This estimator respects know population totals while adjusting the initial design weights as little as possible. Our implementation of this method relies on auxiliary information from Statistics Denmark’s registers taking into account different sampling probabilities and differential non-response.
The weighting was done at household level, where responding households were calibrated to fit population totals for the household population. The weights align the household population with respect to:
The sum of income in each of seven equivalised disposable income intervals
The number of households at risk/not at risk of poverty in each of the five regions (NUTS2)
The number of households at risk/not at risk of poverty in four socioeconomic groups (defined by the main income holder in the household)
“Risk of powerty” is operationalized as equivalised disposable income below 60 pct. of the median.
In addition to correcting for bias due to non-response on the household level, the SILC cross-sectional weights should reproduce certain demographic and poverty distributions on the personal level. Therefore the weights align the population of individuals with respect to:
The number of persons in seven age-groups (0-15, 16-24, 25-39, 40-49, 50-59, 60-69, 70+) living in households at risk/not at risk of poverty
The number of persons of each gender in two age-groups (0-39, 40+) living in each of the five regions
The number of persons living in households below and above the median of equivalised disposable income
Additionally the number of persons 16 years or older are aligned for each gender in five age-groups (16-29, 30-44, 45-54, 55-64, 65+). For this last requirement the categorization according to age and gender for all persons in a responding household is determined by the categorization for the contact person.
Prior to 2016 NUTS2 were not part of the weighting procedure.
From 2011 the income mass within income groups was part of the weighting procedure. This has been done in order to obtain better consistency between our register data and the EU-SILC data and to reduce standard deviations.
18.5.3. Estimation and imputation
Gross values for incomes stem from registers collected from tax authorities, meaning that everyone having an income in Denmark the previous year are covered. Therefore there are no imputed incomes in the gross incomes. There are some smaller parts of the taxes paid (HY140G) which rely on assumptions, due to differences between the Danish tax system and SILC-definitions of taxes paid.
Dividends and sales of stocks are taxed collectively, therefore the tax rate on dividends is assumed to be equal to the average tax-rate paid on income from stocks overall. The taxes on gains and losses has thus been excluded.
People with interest expenses get a tax credit that lowers the overall taxes. As interest payments are not subtracted when calculating the net income, it would be misleading to include the tax credit in the taxes paid. The actual rate varies with income and is therefore modeled.
Profits from self-employment can be pushed into the future, thereby postponing large parts of the tax payment. This leads to large negative net incomes in the taxation year, as the income is generated in one year and taxed in another. This has been sought mitigated by modeling and adding tax payments in the year the income is generated and subtracting tax payments in the year the tax have been paid.
The European Union Statistics on Income and Living Conditions (EU-SILC) is a survey-based instrument aiming at collecting timely and comparable cross-sectional and longitudinal multidimensional microdata on income, poverty, social exclusion and living conditions. In addition, it collects module variables every three years, six years or ad-hoc new policy needs modules.
The EU-SILC instrument provides two types of data:
Cross-sectional data pertaining to a given time or a certain time period with variables on income, poverty, social exclusion and other living conditions;
Longitudinal data pertaining to individual-level changes over time, observed periodically over four‐or more year rotation scheme (Annex III (2) of 2019/1700). Social exclusion and housing condition information is collected mainly at household level while labour, education and health information is obtained for persons aged 16 and over. The core of the instrument is income information at very detailed component level and mainly collected at personal level.
Further details are provided in items 5, 15.1.1.1, 15.2.2 and 18.3.
The source or procedure used for the collection of income variables
The Danish data on incomes are mainly collected in gross form from registers, stemming from the Danish Tax Authoritie. The only exception is intra-household transfers, that are collected via survey.
Many variables are therefore partly based on Danish tax regulation by design. For this reason, a number of adjustments are made to the variables in order to align the data to the SILC definition of incomes. This is possible for most variables, see a more thorough description on the publication 'revision of SILC incomes'.
The tax paid is subtracted from the gross household income to produce the net household income. Because the Danish tax system is complex and almost all income and deductions are part of a integrated tax calculation, it is not possible to calculate person tax rates for all net income components. The total household income is then distributed among the net components the following way. All income components that are exempt from tax are subtracted from the gross income. Tax on pensions from indicidual private plans are taxed separately, this tax is removed from both the gross income and the tax paid, to be added later. The adjusted tax paid and gross income are then used to calculate an effective tax rate at the person level. The tax rate is fixed between 0 and 100 percent of the gross income, any tax outside of this range is attributed to income components according to where the largest gross incomes are found. Typically negative tax rates and tax rates above 100 are found for self-employed persons, so most of this excess tax is placed in PY050N.
Comparability and deviation from definition for each income component
All income components are fully comparable
Household definition and membership
Reference population - Persons living in private households on Danish territory at the end of the income reference period Private household - A person living alone or a group of persons who live together, providing oneself or themselves with the essentials of living Household membership - A person residing in the household or being temporarily away from the household and sharing the cost of living to a large degree
Description of reference period used for incomes
Period for taxes on income and social insurance contributions: 01 January 2023 - 31 December 2023 Income reference periods used: 2023 Reference period for taxes on wealth: 31 December 2023 Lag between the income ref period and current variables: 2-5 months, depending on the date of the interview. Register data for March are typically used for current register variables
Statistical units are private households and all persons living in these households who have usual residence in the Member State. Annex II of the Commission implementing regulation (EU) 2019/2242 defines specific statistical units per variable and specifies the, content of the quality reports on the organization of a sample survey in the income and living conditions domain pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council.
There are two types of general statistical units in the SILC survey.
Persons - A person living in a private household at the end of the income reference period or a current household member of the sampled person
Households - The entire household of the sampled person
Specific statistical units per variable are defined in Annex II of the Commission implementing regulation (EU) 2019/2242 specifying content of the quality reports on the organization of a sample survey in the income and living conditions domain pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council.
The target population is private households and all persons composing these households having their usual residence in the Member State. Private household means a person living alone or a group of persons who live together, providing oneself or themselves with the essentials of living.
Denmark
Description of reference period used for incomes
Period for taxes on income and social insurance contributions
Income reference periods used
Reference period for taxes on wealth
Lag between the income ref period and current variables
2023
2023
2023
3-5 months lag depending on interview date
According to Reg. (EU) 2019/1700 Annex II, precision requirements for all data sets are expressed in standard errors and are defined as continuous functions of the actual estimates and of the size of the statistical population in a country or in a NUTS 2 region. For the income and living conditions domain, the estimated standard errors of the following indicators are examined according to certain parameters set:
· Ratio at‐risk‐of‐poverty or social exclusion to population;
· Ratio of at‐persistent‐risk‐of‐poverty over four years to population;
· Ratio at‐risk‐of‐poverty or social exclusion to population in each NUTS 2 region.
Further information is provided in section 13.2 Sampling error.
Possible sources of bias
Population definition - Some bias may arise when we remove households with 10 adults or more.
Contact at household - Some households cannot be located. If those households differ from the households that can be contacted, a bias will arise
Non-response bias - The persons who refuse to participate in the survey are not representative of the sample as a whole. Typically, younger individuals and individuals with lower incomes are more inclined to refuse to participate in the SILC survey.
These problems are partly mitigated through the calibration of weights (see section 18.5). The calibration ensures that AROP60 almost exactly matches the register, so variables that are closely related to the AROP60 have a lower selection bias than variables that are not that related to AROP60.
Steps are continuously taken in order to reduce the non-response rate, which is the largest source of bias effecting the results.
The data involves several units of measure depending upon the variables. Income variables are transmitted to Eurostat in national currency. For more information, see methodological guidelines and description of EU-SILC target variables available on CIRCABC
Data compilation is carried out in several steps, as described below. The first steps of the process only pertain to cross-sectional data. Later, data from previous years are added and longitudinal weights are calculated.
Raw data manipulation and sample update
Data are received from our data collection partner via a secure FTP-server. Data is then transformed to be more compatible with SILC data formats and duplicates are removed. The sample is updated with metadata on the interview and all persons are assigned personal id's (PB030/RB030). The consolidated sample is then used to create the D, H, R and P populations.
Survey and register variables
The data from the survey is then extracted and recoded to comply with modalities in EU-SILC regulation. Seperate programs extract data from registers and edit to comply as well.
Collection of data, weighting and flag coding
When all variables are produced, they are collected into the D, H, R and P files. Flags are then coded according to regulation and added to the respective files. Weights are then calculated on both a personal and household level.
Longitudinal data
When the cross-sectional data is produced, the data from tre previous three years are gathered from last years transmission. This collected data is used to produce longitudinal weights and ultimately the final EU-SILC dataset.
Data on household composition, current labour market participation and subjective questions are based on interviews. Objective data on housing, education, basic demographics and incomes are based on administrative registers where available.
Survey data
The variables that either cannot or may not be collected from registers are collected via a survey. The survey is designed to meet the definitions laid out in the regulation governing each survey year, customized to fit a national context. The data collection is carried out in spring, by telephone and web.
Register data
A large part of the SILC survey relies on register data. In general, if a register exists that corresponds with a EU-SILC variable, the register is used instead of adding said variable to the survey. The majority of variables related to income, wealth, population and the dwelling are covered by register. These are the main registers used in the compilation of the SILC variables.
Population register (CPR) - Centrally administered population register containing information on the Danish population, with information on sex, age, id's of fathers, mothers and possible spouses, home addresses etc.
Income register - Register of incomes from the Danish tax authority covering all persons paying tax in Denmark
Wealth register - Register of wealth, collected from numerous sources. Covers most financial and non-financial assets and liabilities of Danish individuals and households
Education register - Data from administrative records of the edicational level and activities of the population
Building register - A register of residential and non-residential buildings in Denmark, containing details on the use of the building and the building itself
e-Income register - A register of incomes sourced from the Danish tax authority. Also contains hours worked and taxable benefits
Various government benefit registers - Numerous smaller registers containing information on social benefits have been used in the compilation
Some information from other registers have been used, but to a lesser degree.
Annual
The micro data were submitted to Eurostat by the end of the survey year and less than 12 months after the end of the income reference period.
Data is generally comparable within Europe. But caution is advised, when interpreting minor differences on indicators between countries due to statististical errors and the difficulties related to cross-border income comparisons.
Denmark has participated in the EU-SILC since 2003.
In 2021, the SILC survey was aligned to the EU Regulation 2019/1700 (IESS Regulation - find more information legal framework). This has introduced some new variables and resulted in changes to some other variables. If there is a breaking change, the variable has received a new name. Therefore there are no known data breaks due to the IESS Regulation.
In 2020, there was a major revision in the method behind the calculation of income variables. Read the detailed description revision of SILC incomes.
In 2018, there was a error in the first days of the data collection. This resulted in 405 rejected households due to a looping issue that caused some household members not to be interviewed. 276 of these households was included in SILC. The labour markets status of the missing houshold members was imuted via income data from the preliminary income register and answers provided in previous years. The issue has been described in more detail in an annex to the quality report.
In 2017, parental leave payments was moved from PY120 to HY050
In 2013, a policy change on private lump-sum pensions, has lead to an increase in private pension pay-outs. The effect is temporary only. As SILC 2014 contains incomes from 2013, the SILC-2014 is the first year affected by this policy-change. The expected effect on gini is
SILC-2013: +0
SILC-2014: +0.31
SILC-2015: +0.25
SILC-2016 and forth: ~-0.1 (Provided that we are unable to impute the lump-sum pensions).
Threshold indicators such as Risk of Poverty should be virtually unaffected as most people with private lump-sum pensions have incomes above the median.
In 2014 and again in 2016, Statistics Denmark revised the calibration of SILC. The new calibration has been implemented going back to 2011.
From 2011, the income mass within income groups has been included in the calibration to fit the register better. This has been done in order to obtain better consistency between our register data and the EU-SILC data and has significantly lowered the deviations between full register data and the silc data, when measuring average income and the gini-coefficient. Furthermore between 2009 and 2010 the household definition in the callibration changed from adresses to a narrow concept of the family.
For consistent data on the gini-coefficient and similar pure economic indicators pre-2010, it's recommended to use statistics based on Danish register data for the entire population.
