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 Geo Code - 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 EU Vocabularies, Eurostat's metadata server or CIRCABC.
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 for 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.
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 EU regulation 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 EU Regulation 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. 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.
A person has his/her usual residence in Sweden if the person has his/her actual place of residence in the country. The person must be listed in the Swedish population register and be in Sweden for six consecutive months or more. Short-term visits to another country do not shorten the duration of stay in Sweden.
3.6.1. Reference population
Definitions of reference population, household and household membership
Reference population
Private household definition
Household membership
The reference population is all private households and all persons composing these households, having their usual residence in Sweden.
The household definition is in line with the EU-SILC regulation for an absolute majority of households. In 2024, information about households was collected from administrative data instead of interview data for 271 private households. A household from administrative data is defined as a residential household and consists of all persons registered as living at the same dwelling. The information comes from the Swedish Tax Agency. It is possible, but not necessary, that persons in the same residential household share income or household expenses with each other
The standard EU-SILC definition is applied to an absolute majority of household members. For deviations, please see the information in the Private household definition.
3.6.2. Population not covered by the data collection
The sub-populations that are not covered by the data collection include: those who moved out of the country’s territory; or those with no usual residence.
3.7. Reference area
Sweden (the whole country)
3.8. Coverage - Time
Annual data, reference year 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 (year N-1)
2023 (year N-1)
Not applicable
The fieldwork period (January-June 2024). Therefore, the lag is at minimum 1 month and at maximum 6 months.
6.1. Institutional Mandate - legal acts and other agreements
EU regulation (EU) 2019/1700 was published 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 EU regulation (EU) 2019/2180 of 16 December, 2019 specifies the detailed arrangements and content for the quality reports pursuant to EU regulation 2019/1700 and EU regulation 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 EU regulation 557/2013 and EU regulation 223/2009 on European statistics.
7.1. Confidentiality - policy
In the special task of agencies for producing statistics, confidentiality applies according to Chapter 24, Section 8 of the Swedish Public Access to Information and Secrecy Act (2009:400). To protect the data on natural persons or enterprises that is subject to confidentiality, it is ensured that such data cannot be disclosed directly or indirectly in the published statistics.
With regard to personal data – that is, information that can be directly or indirectly attributed to a living person – the Official Statistics Act (2001:99), the Official Statistics Ordinance (2001:100) and the EU General Data Protection Regulation (2016/679) apply. Statistics Sweden’s confidentiality policy (available only in Swedish) can be found at: Sekretesspolicy 2022-11-02 (scb.se)
7.2. Confidentiality - data treatment
The Swedish EU-SILC follows Statistics Sweden’s Confidentiality Policy (Sekretesspolicy 2022-11-02 (scb.se)). Microdata are protected by the Swedish Public Access to Information and Secrecy Act (2009:400). Rules and information on the processing of personal data can be found in Statistics Swedens’s Data Protection Policy (Data protection policy 2021-09-14 (scb.se)).
In line with the Community legal framework and the European Statistics Code of Practice, Eurostat disseminates European statistics on Eurostat's website (see section 10 - 'Accessibility and clarity'), respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably.
Estimates are published once a year in the Statistical Database on Statistics Sweden’s webpage, Statistics on Income and Living Conditions (SILC) (scb.se). The statistical database contains a wide range of indicators derived from both SILC data and national variables. It is available for everyone to use free of charge.
When making a retrieval from the Statistical Database, the statistics are shown in the form of a table that is based on the subject and variables that have been chosen. For instance, you can choose to show statistics from a certain region, age group or certain years. The tables can also be downloaded as statistics files in different formats.
10.3.1. Data tables - consultations
Not available.
10.4. Dissemination format - microdata access
Researchers or persons affiliated with Swedish organizations can access anonymised microdata following a confidentiality assessment, provided that Statistics Sweden considers that the applicant is eligible to process the data. Information on how to order microdata can be found here: Ordering microdata (scb.se)
10.5. Dissemination format - other
In addition to the statistics published at Statistics on Income and Living Conditions (SILC) (scb.se), users can also order additional data related to variables or study domains that are not included in the official publication at the website.
10.5.1. Metadata - consultations
Not available.
10.6. Documentation on methodology
The annual national quality report and other methodological documents are available in Swedish at:
The overall guidelines for Statistics Sweden’s quality management are described in Statistics Sweden’s quality policy. Quality Policy, 1 March 2022 (scb.se)
Information on data quality, based on the standard quality criteria, is reported in annual national quality reports, available (in Swedish) at the statistics of Sweden website.
Below is also a summary compilation based on standard quality criteria. We cross reference to sections where the results are presented in more detail.
European Statistics on Income and Living Conditions (SILC) are mainly used to present international comparisons regarding income and living conditions in Sweden and other countries within the EU as well as the countries Norway, Switzerland, Iceland, Northern Macedonia, Serbia, and Turkey. SILC is primarily used for monitoring income and living conditions by the European Commission and by researchers from various European countries. A description of e.g., main characteristics, target population and classifications can be found in section 3.
The sources of uncertainty that are assessed to have the greatest impact on key estimates in the Swedish EU-SILC are sampling (due to a specific sample having been studied), non-response (due to answers being missing completely or partly for individuals in the survey) and measurement (e.g., due to misunderstanding of questions or answers). This is further described in subsection 13.1.
The first delivery to Eurostat took place in December 2024, i.e. approximately 6 months after the end of the data collection. For further description of dates of the dissemination of national results, see section 14.
Cross-sectional estimates, from 2008 and onwards, are generally considered to be comparable over time. From 2021 a calibration approach is used to compute longitudinal weights. Comparisons over time of, e.g., PAROP should therefore be made with great caution. The calibration approach is explained in section 18.5.
Comparisons between Swedish SILC and national accounts is provided in Annex 7.
12.1. Relevance - User Needs
The main users of EU-SILC statistical data are: policy makers, government agencies, universities, research institutes, media, and students.
A User Council for Living Conditions and Gender Statistics is held two times per year. The purpose is to create a network of organized user contacts through which Statistics Sweden can gain knowledge about new statistical needs as well as anchor changes with key stakeholders, such as policy makers, government agencies and researchers. The User Council has an active role in questions related to priorities and assessments as well as serving as an important advisory function to Statistics Sweden.
12.2. Relevance - User Satisfaction
Eurostat carried out an online general User Satisfaction Survey (USS) in the period between April and July 2019 (repeated in June-July 2022) to obtain better understanding about users’ 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 microdata 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. Users emphasized their strong need for more detailed microdata.
Variable PB265 (Nature of participation in the survey) has not been transmitted to Eurostat. Unless the questionnaire is answered by the selected respondent, Statistics Sweden does not collect information regarding the personal ID of the person who answers the questionnaire. Hence, the information required to derive PB265 is not available.
HY145 Repayments/receipts for tax adjustment is not collected. Since the income at component level is reported gross, adjustments are instead recorded in the variable HY140G, in accordance with the regulations set out for SILC 2024.
HY170 Value of goods produced for own consumption is not collected. Based on an assessment made by the Swedish Household Budget Survey (HBS), the value of goods produced for own consumption is not considered to constitute a significant component of income.
The following optional variables were not collected in 2024: RL080, HI130G and HI140G
12.3.1. Data completeness - rate
Not available.
13.1. Accuracy - overall
This quality component is related to the closeness of estimates to the true values and its components are variance and bias. Sampling and non-sampling errors were evaluated to define the sources of uncertainty that are assessed to be of the greatest significance to the survey. The main sources of uncertainty in the Swedish EU-SILC, as regards the impact on key estimates, are assessed to arise from (in order of magnitude) sampling, non-response, and measurement:
Sampling: a random uncertainty arises because the survey is based on a sample.
Non-response: non-response occurs when the value of one or more variables in a survey cannot be collected. If all the values for an observation unit are missing, it is called unit non-response. If only some of the values are missing, it is a question of item non-response. To reduce the distortion effects of unit non-response, we use auxiliary information in the estimation process. The estimation process is described in subsection 18.5.
Measurement: measurement errors can have several sources. Thus, for the survey year 2024, measurement errors could arise from two different modes of collection in the Swedish EU-SILC. To reduce errors in CATI, for example due to interviewer effects, co-listening is applied in interviews. For both CATI and CAWI, measurement errors can occur due to memory errors or misunderstanding of questions by the respondent. No cognitive study has been conducted to quantify measurement error sizes.
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.
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 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 the 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 from 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 three 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.
For the income components, the mean, the total number of observations (before and after imputation) and the standard error is reported in Annex 3.
Standard errors in Annex 3 were calculated using PROC SURVEYMEANS SAS procedure without cluster specification. The design was assimilated to a one stage stratified type. A variable available in the national dataset was taken for strata specification, DB090 and PB040 were chosen as weights.
Note that all the standard errors published by Statistics Sweden as well as standard errors of the main indicators used with the quality assessment and presented in Annex A 13.2.1 were calculated in line with the national framework with stratification, clustering as well as household size taken into account. Calculations were made using SAS macro ETOS (Estimation of Totals and Order Statistics) which was designed to compute point and standard error estimates of totals and order statistics (parameters) from sample surveys as well as rational functions of these parameters. According to ETOS 2.0 User’s guide (2012) the Estimating Equations (EE) technique was used for estimation of the variance of the order statistics, and the Taylor linearization method was used for the variance estimation of non-linear functions like ratios and products.
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 the 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.
The main indicators, standard errors and confidence intervals provided in Annex A are calculated by Statistics Sweden, as explained in section 13.2.
The precision requirements defined for the indicator “Ratio of at‐risk‐of‐poverty or social exclusion to population were met, both at a national level and in each of the NUTS2-region level.
The precision requirement related to the indicator “Ratio of at‐persistent‐risk‐of‐poverty over four years to population” at the national level was not met. The national estimate of the standard error of the indicator at the national level was 0,8 versus its respective estimated threshold 0,7. The experiment conducted in 2022 (the control group (CATI only) was excluded from all the samples to prevent breaks in series). This could be mentioned as a possible reason that might have affected the precision of the variance estimate.
13.3. Non-sampling error
Non-sampling errors are basically of four types:
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 the absence of information for the whole units (households and/orpersons) 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.
Of these four types of non-sampling errors in the Swedish EU-SILC, we believe that non-response errors have a greater impact on key estimates than the other three types. See also subsection 13.1 for an enumeration of central sources of uncertainty.
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. As stated in 18.1.3, the Total Population Register (TPR) is used as the sampling frame. The over-coverage in TPR mainly consists of delays in reporting deaths and emigration. People who have left Sweden but are still registered as Swedish residents are difficult to discover but the estimated error is negligible.
Under-coverage: refers to units not included in the sampling frame. The under-coverage in TPR consists mainly of immigrants (immigrants + returnees) that are added to TPR with a certain lag.
Misclassification: refers to incorrect classification of units that belong to the target population.
The coverage deficiencies in the TPR are estimated to be small.
13.3.1.1. Over-coverage - rate
Coverage error
Main problems
Population (sub-population)
Size of error
Comments
Over-coverage
Estimated total over-coverage in TPR
77 400 individuals
Less than one percent of the TPR-population
Under-coverage
Estimated total under-coverage in TPR
23 500 individuals
Misclassification
13.3.1.2. Common units - proportion
Optional
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 caution should be observed in the interpretation of responses to questions related to attitudes and frequency. Most of the questions in the survey refers to the present for which memory errors is not a major source but there are questions about frequency during a longer reference period that are more complicated. For more information regarding measurement errors see section 13.1.
Indirect (proxy) interviews can be a source of errors. In 2024, the proxy interview rate was 1.3 percent.
In 2022, the data collection method in the Swedish SILC was changed from CATI to mixed mode, combining CAWI and CATI. Therefore, from 2022 there is both a telephone questionnaire and a web questionnaire. Both questionnaires have been constructed in cooperation with the cognitive measurement unit at Statistics Sweden.
Some variables are collected via two questions in the phone questionnaire but only one question in the web survey.
This is done to reduce the number of questions in the web survey, where the answer options are visible to the respondent. Furthermore, almost no question in the web survey is mandatory, i.e. it is possible to skip/ not answer a question but still continue the survey. Also, to avoid creating a situation where a respondent chooses to cancel the survey, there are very few direct built-in controls in the web survey. In the phone questionnaire, some controls are still carried out if for example an answer seems to be unreasonable (e.g. extremely high or low).
The questionnaires follow the rules and suggestions made by Eurostat.
Following a basic introductory two-day course about interviewing technique and training in the CATI system, new interviewers participate in an additional three-day course about the EU-SILC and the Swedish living conditions survey. This three-day course includes the following:
• Lectures on different parts of the questionnaire, with time for questions and discussions.
• Test interviews, where the new interviewers conduct telephone interviews with test respondents who are personnel who know EU-SILC well. The new interviewers receive feedback on their test interviews and get the opportunity to ask questions in a two-person situation rather than in a group situation.
• Self-studies, where the new interviewers answer questions about EU-SILC (based on what they have learned during the lectures and what they can read in the document with interviewers’ instructions about EU-SILC) and receive feedback on their answers.
Experienced interviewers, who have worked with EU-SILC during previous years, participate in a one-day course, which begins with self-study of the questionnaire and interviewers’ instructions as well as working with study questions. The one-day course also includes a lecture about new parts and changes in the questionnaire compared to previous year, as well as a group discussion about the study questions.
During the data collection all interviewers have access to mentors, who are experienced interviewers, if they have questions or need help. Interviewers also contact personnel who work with the questionnaire of EU-SILC when they have questions about the content based on, for example, a specific respondent’s situation and/or answers. Interviewers also receive feedback at least once per year from colleagues who are monitoring EU-SILC telephone interviews.
New questions are tested by the cognitive measurement unit at Statistics Sweden.
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: unit non-response and item-non response.
1) Unit non-response which refers to the absence of information of the whole units (households and/or persons) selected into the sample.
To reduce the bias due to non-response, we use auxiliary information in the estimation process. The estimation process is described in Annex 5 – Weighting procedure.
According to Annex VI of the EU regulation 2019/2242:
Household non-response rates (NRh) are computed as follows:
NRh=(1-(Ra * Rh)) * 100 Where Ra is the address contact rate defined as: Ra= Number of addresses/selected person (including phone, mail if applicable) successfully contacted/number of valid addresses/selected persons (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 the database/number of eligible households at contacted addresses (including phone, mail if applicable).
Individual non-response rates (NRp) are 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 interviews completed/number of eligible individuals in the households whose interviews were completed and accepted for the database.
Overall individual non-response rates (*NRp) are computed as follows: *NRp=(1-(Ra * Rh * Rp)) * 100
For Sweden, where a sample of persons rather than a sample of households (addresses) was selected, the individual non-response rates will be calculated for ‘the selected respondents'.
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
95,92
94,68
96,40
54,13
56,95
55,48
100
100
100
48,08
46,08
46,51
0
0
0
48,08
46,08
46,51
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.
* A change in the coding of DB120 and DB130 between 2023 and 2024 has resulted in a notable increase in the rate of respondents coded as DB120=11 for SILC 2024 compared to SILC 2023. The increase is not due to an observed change in response patterns.
13.3.3.2. Item non-response - rate
The computation of item non-response is essential to fulfil the precision requirements. The item non-response rate is provided for the main income variables, both at household and personal level. Item non-response 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
Annex 2 contains the following information:
Percentages of households/persons having received an amount for each income component.
Percentage of missing values for each income component collected or compiled at household/personal level
Percentage of partial information for each income component collected or compiled at household/personal level.
13.3.4. Processing error
Description of data entry, coding controls and the editing system
Data entry and coding (if any used)
Editing controls
Data are checked interactively (value, syntax, logic) as an integrated part of the data entry process.
See section 18.4 Data validation.
13.3.5. Model assumption error
A model assisted estimation approach is used. Calibration is used to reduce the non response error. The calibration approach is further explained in Annex 5.
14.1. Timeliness
The data collection took place during January-June 2024. Cross-sectional and longitudinal target variables, including cross-sectional and longitudinal weights, were submitted to Eurostat on 20 December 2024, i.e. at the end of reference year 2024.
14.1.1. Time lag - first result
Only final results are published. See section 14.1.2
14.1.2. Time lag - final result
Final results were published on 21 February 2025, i.e. one month and three weeks after the end of reference year 2024. Statistics Sweden’s publishing calendar is available at: Publishing calendar (scb.se)
14.2. Punctuality
There was no time lag between the actual delivery of the data and the target date when it should have been delivered according to the transmission deadlines.
14.2.1. Punctuality - delivery and publication
For punctuality on the delivery of the data see section 14.2.
Final results were published in February 2025 (one month and three weeks after the end of reference year 2024), in accordance with the current schedule, which makes the percentage of data release delivered on time 100 percent.
15.1. Comparability - geographical
The data is comparable between NUTS2 regions.
15.1.1. Asymmetry for mirror flow statistics - coefficient
Not applicable
15.2. Comparability - over time
There have been no significant changes in the design between 2008 to 2020. The comparability between those years is therefore considered to be good. Comparisons between estimates before and after 2008 should be made with great caution. This is mainly due to two reasons. One is that from year 2007-, the data collection is carried out mainly through CATI. In 2004 CAPI was mainly used and in 2006 about half of the interviews were conducted through CAPI and about half were conducted with CATI. The second reason is that from 2016, a calibration approach is used to calculate cross-sectional weights. In 2016, the cross-sectional weights for 2008 to 2015 were recalculated with the calibration approach. In 2021 a new longitudinal calibration estimation procedure was implemented which in turn might have affected the comparability with previous years but for only longitudinal estimates.
A review of the national questionnaire is made each year to ensure that the content complies with existing directives regarding EU-SILC and that the respondent perceives the questionnaire to be clear and intelligible. In some instances, this can lead to breaks in series for some variables.
Variables related to childcare (i.e. RL-variables) were reviewed and reinterpreted before the 2019 data collection. In 2018, preschool class became compulsory in Sweden, and children in preschool class are thus classified in RL020 together with students in primary school since 2019. Before 2019, children in preschool class were classified in RL010. Children at day-care centres were recoded from RL040 to RL010 and some of the children that get childcare by a professional child-minder were recoded from RL050 to RL040. These changes resulted in breaks in series and time comparisons with previous years should thus be avoided.
From 2021, questions regarding the Global Activity Limitation Instrument (GALI) are implemented in all surveys covered by the new EU framework regulation on social statistics. To ensure that the directives from Eurostat regarding GALI are followed, a common design of the questions regarding GALI are now implemented in the relevant surveys conducted by Statistics Sweden. For SILC, GALI is collected via two questions instead of four since 2021.
These changes resulted in breaks in series and time comparisons with previous years should thus be avoided.
In 2022 the data collection method in the Swedish SILC was changed from CATI to mixed mode combining CAWI and CATI. This means that the respondents from 2022 and onwards may choose if they want to respond via telephone interview, or via a web-based questionnaire. In order to evaluate the effects of the change in data collection method, Statistics Sweden conducted a split-sample experiment where the results from the mixed mode data collection were compared with results from a parallell data collection where only telephone interviews were used. The comparison between the control group (CATI only) and the experiment group (CAWI and CATI) showed that some of the SILC variables have been affected by the change in data collection method. Variables for which a statistically significant difference is observed (p-value < 0.05) and where further analyses also indicate a break in series, are listed in column F of Annex 8 for the year 2022. It is recommended to avoid any time comparisons with previous years if the listed variables are involved.
Changes that have taken place in 2024 are described in Annex 8.
15.2.1. Length of comparable time series
2008 - 2020: As described in subsection 15.2, there have been no significant changes in the design between 2008 to 2020. The comparability between those years is therefore considered to be good.
2021: In 2021 it was recommended to avoid time comparisons for longitudinal estimates with any previous year's values due to the new longitudinal calibration estimation procedure.
2022: In 2022 It is recommended to avoid any time comparisons with previous years if the variables listed in column F of Annex 8 for the year 2022 are involved.
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)
F
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)
L
An estimated value based on register data for various interest payments.
Income received by people aged under 16
(HY110)
F
Regular taxes on wealth
(HY120)
F
Taxes paid on ownership of household main dwelling
(HY121)
L
More than one dwelling/property might be included
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
Value of goods produced for own consumption
(HY170)
NC
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)
F
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.
A comparison with the Household Budget Survey (HBS) is not available.
15.3.1. Coherence - sub annual and annual statistics
Not applicable
15.3.2. Coherence - National Accounts
See Annex 7 – Coherence
15.4. Coherence - internal
There is no clear lack of coherence in the EU-SILC data set.
The mean interview duration for selected respondents is calculated as the sum of the duration of all personal interviews (sum of HB100), divided by the number of questionnaires completed for selected respondents. For SE, HB100 is the total interview time with the selected respondent. In 2024. for interviewing mode CATI, the mean (average) interview duration for direct interviews with selected respondents is 27 minutes. For telephone interviews where a professional translator is used the mean duration is 52 minutes. For interviews where the interviewer translates parts of or the whole interview to the selected respondent’s language, the mean duration is 35 minutes. For CAWI, the mean duration is 23 minutes.
17.1. Data revision - policy
Statistics Sweden’s overall data revision policy can be found here: Revideringspolicy på engelska, 2022 (scb.se). There is no specific data revision policy related to the Swedish EU-SILC.
17.2. Data revision - practice
In 2016, Statistics Sweden developed and implemented new cross-sectional weights for SILC data. Consequently, previously published estimates were replaced with estimates based on the new cross-sectional weights for the years 2008 to 2015.
17.2.1. Data revision - average size
See section 17.1 and 17.2.
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
The source of data is a combination of data collected through interviews, and registers.
Raw data is collected by computer-assisted telephone interview (CATI, in exceptional cases computer assisted personal interview (CAPI)) and computer assisted web-interview (CAWI). Data is then supplemented with data from administrative sources/registers. If the selected respondent is unable to respond, a CATI or CAWI-proxy interview can be carried out with either a member of the household or a person outside of the household, chosen by the selected respondent. If needed, respondents are interviewed by phone using a professional translator.
The following administrative registers are used:
The Longitudinal integrated database for health insurance and labour market statistics (LISA): The database comprises detailed data on health insurance, parental benefit, and unemployment benefit at the individual level. LISA enables the study of individuals’ transition over time between, for instance, gainful employment, unemployment, and illness. In EU-SILC, LISA is used to compute variables on labour market participation.
The Labour statistics based on administrative sources (RAMS): The statistics show employment, commuting, the composition of personnel and the industrial structure. They also show events and flows in the labour market. The statistics are complete and can be broken down to a low regional level or based on the employees' characteristics, for example, sex, education and age. In EU-SILC, RAMS is used to compute variables on labour market participation.
The monthly employer reports at individual level (AGI): The statistics show payments and tax deductions for each payee each month. All registered employers (e.g. companies or associations) are obliged to submit this information to the Swedish Tax Agency on a monthly basis. In EU-SILC, the AGI is used to compute variables on labour market participation.
The Population and housing census 1960–1990 (FoB): Statistics show different aspects of society at the time of the census. There are statistics available from FoB-75 to FoB-90 on the population's employment, household composition and accommodation. In EU-SILC, FoB is used to compute variables on labour market participation.
The Register on Participation in Education (UREG): UREG is an individual-level register, which serves as the basis for statistics on the educational attainment of the population. Information on completed education is continuously reported to Statistics Sweden by the country's schools and education providers, and annually entered in UREG. UREG measures a person’s highest completed level of education up to, and including, the spring semester before the current turn of the year. In EU-SILC, UREG is used to compute variables on educational attainment and background.
The Total Population Register (TPR): The TPR contains information about the population and its changes, and it reflects largely the content of the population register of the Swedish Tax Agency. It is the foundation of official population and household statistics. Examples of population statistics include population by sex, age, marital status etc. in counties and municipalities. For more information about the TPR, se section 18.1.3. In EU-SILC the TPR is mainly used to compute standardised and core variables (i.e. variables starting with DB, HB, RB and PB).
The Income and Taxation Register (IoT): The IoT is an individual-level register and serves as the basis for the official income and tax statistics. The register is produced once a year, and the reference period is each income year. The register includes all taxpayers and registered persons, as well as estates. The production of the IoT consists of the collection and processing of data from administrative sources into a final observation register. The register contains information from the following government agencies: the Swedish Tax Agency, the Swedish Social Insurance Agency, the Swedish Board of Student Finance, the National Government Employee Pensions Board, the Swedish Pensions Agency, the Swedish Armed Forces, the Swedish National Agency for Education and the National Board of Health and Welfare. The register is also supplemented with information from the TPR. In EU-SILC, the IoT is mainly used to compute variables related to income (i.e. variables starting with PY and HY).
The Real Property Register: The Real Property Register is an administrative register that contains information about all properties, buildings, addresses and apartments in Sweden. The register is managed by the National Land Survey, which also is the responsible agency. The register is updated regularly via weekly notifications from the National Land Survey. Year-end versions have been saved since 2015. The Real Property Register also includes the Dwelling Register, a national register of all dwellings in Sweden. In EU-SILC, the Real Property register is used to compute variables on person and household characteristics and main housing characteristics.
The Property Assessment Register (FTR): The primary purpose of the FTR is to determine assessment values for taxable properties. The statistics annually report the outcome of general, simplified and special property assessments. In EU-SILC, FTR is used to compute the variable imputed rent (HY030G), which is included every 3 years starting in 2023.
18.1.1. Sampling Design
The Swedish SILC utilizes the selected respondent model; hence, the description below refers to the sample of selected respondents. Each selected respondent provides information about household members. The response set of selected respondents and household members is hence the response set of a network sample of selected respondents and household members.
The Swedish SILC is a rotating panel survey. Until 2020, the survey had four rotating panels. In 2021, the number of panels increased to five, and in 2022, the number of panels increased to six, which is the number of panels from 2022 and onwards. All panels, which were part of the 2020 sample, will participate in the survey for six years, which facilitates the transition between four and six panels.
The actual cross-sectional sample size of the Swedish SILC is 20 000 selected respondents. Up until 2020, the cross-sectional sample size was 11 600 selected respondents, and the panel sample size was 2 900.
When the survey has six rotating panels with similar sample size, they will each have a sample size of 3 335 selected respondents. New panel samples from 2021 will hence have a sample size of 3 335. To reach the desired cross-sectional sample size, there will be additional yearly cross-sectional samples 2021-2023. Because the panel from 2020 will be in the cross-sectional sample until 2025, the new sample size structure will be fully implemented from 2026.
In 2022, a split-sample experiment was conducted to examine time series breaks because of the introduction of mixed mode data collection. Estimates from this survey round were only based on the respondent group that were exposed to mixed mode data collection in the experiment. Thus, sample and panel sizes differ in 2022 from the general pattern.
The sampling design for panel samples selected prior to 2021 was one-stage stratified simple random sampling, where stratification was with respect to ages 16-24 years, 25-34 years, 35-44 years, 45-54 years, 55-64 years, 65-74 years, 75-84 years, and 85 years and older. The sample allocation was proportional to the stratum size. From 2021 and onwards, the sampling design for new panel samples is one-stage stratified systematic sampling, in which stratification is with respect to NUTS2 regions and where the sampling frame is ordered by sex and age. The sampling allocation is non-proportional, in which small NUTS2 regions are overrepresented, and large NUTS2 regions are underrepresented. The purpose of the new sampling design is to increase compliance with the precision requirements for SILC.
The sampling design for the additional samples 2021-2023 is stratified systematic sampling, for which the stratification, allocation, and ordering of the sampling frame is the same as for new panel samples. The sampling allocation for the additional samples is more skewed relative to the population than the sampling allocation for new panel samples, which further increases compliance with the precision requirements.
Each year, supplementary samples of individuals, which have become eligible for participation in the survey during the previous year, either from age, i.e., by turning 16 years old, or by immigration, are selected for all panels, which were part of the survey during the previous survey round. The sample size is proportional to the size of the group of newly eligible individuals relative to the population. The sampling design for the supplementary samples correspond to the sampling design for the panels for which they are selected. This ensures that panel samples stay representative of the current cross-sectional population.
For information on achieved sample size for the year 2024, and rotational group breakdown, please see annex 4, table 2.
18.1.2. Sampling unit
The sampling unit is individuals (selected respondents).
18.1.3. Sampling frame
The Total Population Register (TPR) is kept at Statistics Sweden since 1968. TPR is an extract from the population register at the Tax Authorities and all individuals residing in Sweden shall be registered at the property unit in the parish where they live. Each individual in TPR has a unique personal identity number. TPR receives daily updates on births, deaths, changes in marital status, and changes in citizenship, national migration, immigration and emigration from the Tax Authorities. Received information is checked mechanically with respect to the validity of the codes and the logical contents of the information and quality tests comprises, among other things, regional codes, connections between age and marital status, etc.
TPR is used as the sampling frame for the selection of the sample of selected respondents. Data refers to the third quarter of the precedent year of the survey year. Persons aged less than 16 years of age are excluded from the frame.
18.2. Frequency of data collection
Annually.
The fieldwork was carried out from January to June 2024.
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
33,1
65,7
0,7
0,6
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
The income variables as well as wealth and taxes are collected by administrative databases and registers at The Swedish Tax Agency and Statistics Sweden.
Gross but exclusive of employers’ social contribution.
The components are gross and available from administrative registers except for employees’ social contribution.
Administrative data sources are reviewed prior to and during entry into the database.
Assessment during data collection During the phone interviews, checks are carried out to ensure that the answers are correct. For example, if the respondent provides a response that clearly deviates from what is considered reasonable, a message will appear in the CATI-software, prompting the interviewer to check the response with the respondent. In addition, the statistician responsible for the subject in question reviews the response data during the first few weeks to ensure that the data collection process is functioning as intended.
Assessment of microdata Data are examined at an aggregate level, which for example means that the proportion of yes-answers for a certain question is compared with the outcome of last year. After the interview, response data is sent to the system where coding, e.g. occupational coding, as well as a review of household characteristics is carried out. The data are then reviewed by the statistician responsible for the subject in question. Target variables are also verified using data checks, provided by Eurostat.
Assessment of macrodata Estimates based on survey data are compared with the corresponding values from previous years. To some extent, income data are also compared with totally register-based Income and Tax statistics to evaluate if the estimates are reasonable.
Assessment of data publication Prior to publication, all parts of the material are examined. Statistics Sweden ensures that all tables and diagrams are included and that no tables are empty or contain incomprehensible values, such as internal codes. Headings and explanations of tables and diagrams are also examined.
18.5. Data compilation
The weighting procedure is described in Annex 5.
For information on imputation see Annex 6. For other information on data editing, see section 18.4.
18.5.1. Imputation - rate
Imputation is the process used to assign replacement values for missing, invalid or inconsistent data that have failed edits. This includes automatic and manual imputations; it excludes follow-up with respondents and the corresponding corrections (if applicable). The unweighted imputation rate for a variable is the ratio of the number of imputed values to the total number of values requested for the variable. This metadata concept complements the information provided in subsections 18.5 and 13.3.4.
18.5.2. Calculation of weighting factors and weight adjustments
The weighting procedure is described in Annex 5.
18.5.3. Estimation and imputation
For information on imputation, see Annex 6.
18.6. Adjustment
Not applicable
18.6.1. Seasonal adjustment
Not applicable
For information on the rolling module, see Annex 9.
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.
Statistical units are private households and all persons living in these households who have usual residence in the Member State. Annex II of the EU regulation 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 EU Regulation 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. 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.
A person has his/her usual residence in Sweden if the person has his/her actual place of residence in the country. The person must be listed in the Swedish population register and be in Sweden for six consecutive months or more. Short-term visits to another country do not shorten the duration of stay in Sweden.
Sweden (the whole country)
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 (year N-1)
2023 (year N-1)
Not applicable
The fieldwork period (January-June 2024). Therefore, the lag is at minimum 1 month and at maximum 6 months.
This quality component is related to the closeness of estimates to the true values and its components are variance and bias. Sampling and non-sampling errors were evaluated to define the sources of uncertainty that are assessed to be of the greatest significance to the survey. The main sources of uncertainty in the Swedish EU-SILC, as regards the impact on key estimates, are assessed to arise from (in order of magnitude) sampling, non-response, and measurement:
Sampling: a random uncertainty arises because the survey is based on a sample.
Non-response: non-response occurs when the value of one or more variables in a survey cannot be collected. If all the values for an observation unit are missing, it is called unit non-response. If only some of the values are missing, it is a question of item non-response. To reduce the distortion effects of unit non-response, we use auxiliary information in the estimation process. The estimation process is described in subsection 18.5.
Measurement: measurement errors can have several sources. Thus, for the survey year 2024, measurement errors could arise from two different modes of collection in the Swedish EU-SILC. To reduce errors in CATI, for example due to interviewer effects, co-listening is applied in interviews. For both CATI and CAWI, measurement errors can occur due to memory errors or misunderstanding of questions by the respondent. No cognitive study has been conducted to quantify measurement error sizes.
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.
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.
The weighting procedure is described in Annex 5.
For information on imputation see Annex 6. For other information on data editing, see section 18.4.
The source of data is a combination of data collected through interviews, and registers.
Raw data is collected by computer-assisted telephone interview (CATI, in exceptional cases computer assisted personal interview (CAPI)) and computer assisted web-interview (CAWI). Data is then supplemented with data from administrative sources/registers. If the selected respondent is unable to respond, a CATI or CAWI-proxy interview can be carried out with either a member of the household or a person outside of the household, chosen by the selected respondent. If needed, respondents are interviewed by phone using a professional translator.
The following administrative registers are used:
The Longitudinal integrated database for health insurance and labour market statistics (LISA): The database comprises detailed data on health insurance, parental benefit, and unemployment benefit at the individual level. LISA enables the study of individuals’ transition over time between, for instance, gainful employment, unemployment, and illness. In EU-SILC, LISA is used to compute variables on labour market participation.
The Labour statistics based on administrative sources (RAMS): The statistics show employment, commuting, the composition of personnel and the industrial structure. They also show events and flows in the labour market. The statistics are complete and can be broken down to a low regional level or based on the employees' characteristics, for example, sex, education and age. In EU-SILC, RAMS is used to compute variables on labour market participation.
The monthly employer reports at individual level (AGI): The statistics show payments and tax deductions for each payee each month. All registered employers (e.g. companies or associations) are obliged to submit this information to the Swedish Tax Agency on a monthly basis. In EU-SILC, the AGI is used to compute variables on labour market participation.
The Population and housing census 1960–1990 (FoB): Statistics show different aspects of society at the time of the census. There are statistics available from FoB-75 to FoB-90 on the population's employment, household composition and accommodation. In EU-SILC, FoB is used to compute variables on labour market participation.
The Register on Participation in Education (UREG): UREG is an individual-level register, which serves as the basis for statistics on the educational attainment of the population. Information on completed education is continuously reported to Statistics Sweden by the country's schools and education providers, and annually entered in UREG. UREG measures a person’s highest completed level of education up to, and including, the spring semester before the current turn of the year. In EU-SILC, UREG is used to compute variables on educational attainment and background.
The Total Population Register (TPR): The TPR contains information about the population and its changes, and it reflects largely the content of the population register of the Swedish Tax Agency. It is the foundation of official population and household statistics. Examples of population statistics include population by sex, age, marital status etc. in counties and municipalities. For more information about the TPR, se section 18.1.3. In EU-SILC the TPR is mainly used to compute standardised and core variables (i.e. variables starting with DB, HB, RB and PB).
The Income and Taxation Register (IoT): The IoT is an individual-level register and serves as the basis for the official income and tax statistics. The register is produced once a year, and the reference period is each income year. The register includes all taxpayers and registered persons, as well as estates. The production of the IoT consists of the collection and processing of data from administrative sources into a final observation register. The register contains information from the following government agencies: the Swedish Tax Agency, the Swedish Social Insurance Agency, the Swedish Board of Student Finance, the National Government Employee Pensions Board, the Swedish Pensions Agency, the Swedish Armed Forces, the Swedish National Agency for Education and the National Board of Health and Welfare. The register is also supplemented with information from the TPR. In EU-SILC, the IoT is mainly used to compute variables related to income (i.e. variables starting with PY and HY).
The Real Property Register: The Real Property Register is an administrative register that contains information about all properties, buildings, addresses and apartments in Sweden. The register is managed by the National Land Survey, which also is the responsible agency. The register is updated regularly via weekly notifications from the National Land Survey. Year-end versions have been saved since 2015. The Real Property Register also includes the Dwelling Register, a national register of all dwellings in Sweden. In EU-SILC, the Real Property register is used to compute variables on person and household characteristics and main housing characteristics.
The Property Assessment Register (FTR): The primary purpose of the FTR is to determine assessment values for taxable properties. The statistics annually report the outcome of general, simplified and special property assessments. In EU-SILC, FTR is used to compute the variable imputed rent (HY030G), which is included every 3 years starting in 2023.
Annual.
The data collection took place during January-June 2024. Cross-sectional and longitudinal target variables, including cross-sectional and longitudinal weights, were submitted to Eurostat on 20 December 2024, i.e. at the end of reference year 2024.
The data is comparable between NUTS2 regions.
There have been no significant changes in the design between 2008 to 2020. The comparability between those years is therefore considered to be good. Comparisons between estimates before and after 2008 should be made with great caution. This is mainly due to two reasons. One is that from year 2007-, the data collection is carried out mainly through CATI. In 2004 CAPI was mainly used and in 2006 about half of the interviews were conducted through CAPI and about half were conducted with CATI. The second reason is that from 2016, a calibration approach is used to calculate cross-sectional weights. In 2016, the cross-sectional weights for 2008 to 2015 were recalculated with the calibration approach. In 2021 a new longitudinal calibration estimation procedure was implemented which in turn might have affected the comparability with previous years but for only longitudinal estimates.
A review of the national questionnaire is made each year to ensure that the content complies with existing directives regarding EU-SILC and that the respondent perceives the questionnaire to be clear and intelligible. In some instances, this can lead to breaks in series for some variables.
Variables related to childcare (i.e. RL-variables) were reviewed and reinterpreted before the 2019 data collection. In 2018, preschool class became compulsory in Sweden, and children in preschool class are thus classified in RL020 together with students in primary school since 2019. Before 2019, children in preschool class were classified in RL010. Children at day-care centres were recoded from RL040 to RL010 and some of the children that get childcare by a professional child-minder were recoded from RL050 to RL040. These changes resulted in breaks in series and time comparisons with previous years should thus be avoided.
From 2021, questions regarding the Global Activity Limitation Instrument (GALI) are implemented in all surveys covered by the new EU framework regulation on social statistics. To ensure that the directives from Eurostat regarding GALI are followed, a common design of the questions regarding GALI are now implemented in the relevant surveys conducted by Statistics Sweden. For SILC, GALI is collected via two questions instead of four since 2021.
These changes resulted in breaks in series and time comparisons with previous years should thus be avoided.
In 2022 the data collection method in the Swedish SILC was changed from CATI to mixed mode combining CAWI and CATI. This means that the respondents from 2022 and onwards may choose if they want to respond via telephone interview, or via a web-based questionnaire. In order to evaluate the effects of the change in data collection method, Statistics Sweden conducted a split-sample experiment where the results from the mixed mode data collection were compared with results from a parallell data collection where only telephone interviews were used. The comparison between the control group (CATI only) and the experiment group (CAWI and CATI) showed that some of the SILC variables have been affected by the change in data collection method. Variables for which a statistically significant difference is observed (p-value < 0.05) and where further analyses also indicate a break in series, are listed in column F of Annex 8 for the year 2022. It is recommended to avoid any time comparisons with previous years if the listed variables are involved.
Changes that have taken place in 2024 are described in Annex 8.
