Swedish Migration Agency

Planning Department

Unit for statistics

Slottsgatan 82, 602 22 Norrköping

This metadata information is provided by the authorities of Sweden to ensure compliance with the requirements of Article 4.4 (Dublin statistics) of the Regulation (EC) 862/2007 on Migration and International Protection as amended by the Regulation (EU) 2020/851.

As required by Article 9 of the Regulation (EC) 862/2007, Member States must report to Eurostat on the data sources used, the reasons for the selection of these sources and the effects of the selected data on the quality of the statistics. Member States shall provide Eurostat with all the information necessary to evaluate the quality, comparability and completeness of the statistical information.

The reference document used for assessing the compliance of the concepts and definitions in this metadata file is the document Dublin statistics - Eurostat Technical guidelines to Dublin statistics (see Annex).

The Swedish Migration Agency does not officially present Dublin statistics. The statistics are presented on internal websites and in case of inquiries. 

The Swedish Migrations agency usually uses the same classifications as Eurostat.

Migration and international protection – Dublin statistics

The Swedish Migration Agency does not officially present Dublin statistics. The statistics are presented on internal websites and in case of inquiries. 

The Swedish Migrations agency usually uses the same classifications as Eurostat.

The text of Art.4 of the Regulation (EC) 862/2007 on Migration and international protection refers in general to statistics based on the number of persons and not on the number of applications. However, the article describing the 'Dublin' statistics (Art.4.4) refers to statistics based on the number of requests. Commission services recommend that Art.4.4 (more precisely 4.4.a, 4.4.c and 4.4.d) is interpreted as referring to the number of persons concerned by each request, decision and transfer.

Third country nationals submitting an application of asylum under the Dublin Regulation.

Sweden.

Since 2015.

Not applicable.

Number of requests, decisions, transfers.

Reference year 2022

The compilation of Dublin statistics is based on Article 4.4 of Regulation (EC) 862/2007 of 11 July 2007 which depicts in details the data provisions Sweden are bound to supply Eurostat with.

Not applicable.

See below.

There is no national policy for (not) releasing the Dublin data before a certain time after the end of reference period has passed.

Not applicable.

The main rule is confidentiality, if an external user would like to access classified information for research purposes the request from the external user will be subject to a confidentiality test. A request can be granted providing (and ensuring) that the information can be de-identified.

Statistical confidentiality is guaranteed by law, in chapter 24 article 8 in the Public Access to Information and Secrecy Act (OSL 2009:400) and article 7 in the Public Access to Information and Secrecy Ordinance (OSF 2009:641). The provision refers specifically to secrecy for the protection of individuals.

All staff at Swedish Migration Agency (SMA) sign legal confidentiality commitments stating that they have taken note of the relevant confidentiality provisions. Breaches of confidentiality are punishable by The Swedish Penal Code (1962:7800) in chapter 20 article 3. 

There are physical, technological and organisational provisions in place to protect the security and integrity of the data at SMA. Organisational provision are laid out according the legislation outlined in chapter 24 article 8 OSL and article 7 in OSF. Microdata are protected by absolute confidentiality by above mentioned articles in OSL and OSF. Only authorised staff, and as few as possible, are given access to sensitive data concerning statistical activities. For access to the statistics systems, there is an eligibility model based on process, role and eligibility.  In order to meet the legal requirements of the Data Protection Regulation (GDPR), the possibility of gaining access to the operational views have been severely limited for the majority of the staff at the agency.

No special data treatment done in regards to confidentiality.

Not officially presented at national level.

Not officially presented at national level.

Not officially presented at national level.

Annual data transmission to Eurostat. Monthly national dissemination. 

Not available.

Not available.

Not available.

Not applicable.

Not available.

Not available.

Not applicable.

Not available.

Not available.

Not available.

Not available.

Dublin statistics in Sweden follow the given guidelines given in art. 4.4 of Regulation (EC) 862/2007. However the underlying data in some cases exhibits various aspects of quality issues from events such as faulty registrations to uncertain unit definition to query limitations. 

There is currently a new data warehouse in development with the aim to significantly increase data quality and data retrieval capabilities. Response given in this metadata questionnaire may thus become obsolete in the future as the new data warehouse is being put into production.

• Good quality (extensive quality) - all quality requirements are met according to the standard quality criteria and in line with the Technical Guidelines for the data collection of Dublin statistics (Annex).
• Sufficient quality (acceptable quality) - minimum quality requirements are met: minor issues exist which need to be improved.
• Poor quality (inadequate quality) - the data do not meet the minimum quality requirements: important/many issues exist which affect the quality of the data and which need to be solved.

The statistics collected under Regulation (EC) 862/2007 are used by a wide range of users at national, European and International level to cover various usersʼ needs.

Users of Dublin statistics : Eurostat, EUAA, Department of Justice, national authorities.

Unmet user needs (on statistics) and reasons why these needs cannot be satisfied (at national and international level) : Data in some cases exhibits various aspects of quality issues from events such as faulty registrations to uncertain unit definition to query limitations. 

User satisfaction not collected.

See below.

85%

Completeness is the extent to which all requested Dublin statistics are provided to Eurostat, according to the data provisions foreseen by Articles 4.4. of Regulation (EC) 862/2007 (as amended by Regulation (EU) 2020/851) and by the Technical Guidelines for Dublin statistics (Annex).

• Complete - The statistics are complete: the dataset and all requested breakdowns are available, as foreseen by Regulation (EC) 862/2007 (as amended by Regulation (EU) 2020/851) and the Technical Guidelines for the data collection of Dublin statistics (Annex)
• Incomplete - The statistics are incomplete: the dataset is available but some of the requested breakdowns are not available, as foreseen by Regulation (EC) 862/2007 (as amended by Regulation (EU) 2020/851) and the Technical Guidelines for the data collection of Dublin statistics (Annex)
• Not available - The statistics are not available: the dataset is not available, that is all the requested breakdowns are missing, as foreseen by Regulation (EC) 862/2007 (as amended by Regulation (EU) 2020/851) and the Technical Guidelines for the data collection of Dublin statistics (Annex)
• Not applicable - The statistics are Not applicable: the concept/process is not used or it does not exist in the country

Dublin statistics in Sweden follow the given guidelines given in art. 4.4 of Regulation (EC) 862/2007. However the underlying data in some cases exhibits various aspects of quality issues from events such as faulty registrations to uncertain unit definition to query limitations. 

There is currently a new data warehouse in development with the aim to significantly increase data quality and data retrieval capabilities. Response given in this metadata questionnaire may thus become obsolete in the future as the new data warehouse is being put into production.

Accuracy is the closeness of the data to the exact or true values that the statistics were intended to measure. Reliability is the closeness of the initial provided value to the subsequent (revised) value.

Assessment of the accuracy of the national Dublin statistics, at dataset level

• High = 100 % accuracy. The reported statistics measure accurately the target population, as foreseen by Regulation (EC) 862/2007 (as amended by Regulation (EU) 2020/851) and as defined in the Technical Guidelines for Dublin statistics (Annex)
• Good = 95 - 99 % accuracy. The reported statistics do not cover accurately the target population and a small part of the targeted population is missing, or a small part of non-targeted (erroneous) population is included in the reported statistics; the bias/error in the reported statistics is expected to be small
• Low = 80 - 94 % accuracy. The reported statistics do not cover accurately the target population and a large part of the targeted population is missing, or a large part of non-targeted (erroneous) population is included in the reported statistics; the bias/error in the reported statistics is expected to be large
• Very low = less than 80 % accuracy. The reported statistics do not cover accurately the target population and a very large part of the targeted population is missing, or a very large part of non-targeted (erroneous) population is included in the reported statistics; the bias/error in the reported statistics is expected to be very large

Not applicable.

Not applicable.

See below.

See below.

See the file ‘Proportion of units (persons) reported in the statistics that do not belong to the target populationʼ in annex.

Not applicable.

See concept 13.3.4 Processing Error.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

See below concepts 14.1.1 and 14.1.2.

- Number of days (on average) after the end of the reference period at which the country is able to report the first/preliminary data, at the earliest : 2 days

- Reasons for possible long production of the first/preliminary data and plans to improve the situation : In the event of a database failure. 

- Number of days (on average) after the end of the reference period at which the country is able to report the final data, at the earliest : 90 days

- Reasons for possible long production of the final data and plans to improve the situation : 

100%

100%

See below.

Not applicable.

Assessment of possible sources of lack of comparability between countries, due to discrepancies from Eurostat concepts, definitions or due to differences in methods.

Comparability over time.

Since 2015.

Only differences in definitions according to guidelines.

Variables pertaining to the individual refer to the most recent registration at the date of data extraction, reporting of the same month can thus vary depending on the date of data extraction.

Not applicable.

No internal incoherence.

Regular statistical reporting and revising to EDAMIS estimated to 20 working days per year. This does not include time to implement changes in guidelines in regular reporting. 

Metadata questionnaire estimated to 10 working days involving 2 workers.

Data is revised continuously during the year.

European statistics are compiled according to Eurostat guidelines on the basis of common standards regarding the scope, definitions, units and classifications of the various surveys and data sources. The Swedish Migration Agency continuously revises previously published statistics on changes in historical data in the previous year. In case of major development/efforts that lead to improved quality of historical data, revisions are transmitted.

In the central database, in principle, all statistics relating to information about the applicant and their cases are stored. However, the structure of the database is inadequate from a follow-up perspective. The new Data warehouse that is under development will be based on more logical structures that make it easier to follow up data in a quality-assured way.

See below.

The size of revision gives an idea of the stability of the data while direction is important to understand whether preliminary data tend to overestimate or underestimate the real target population.

Difficult to measure changes after revision.

Swedish Migration Agency.

Daily.

Data is directly registered by operative personnel into the central database. Data is then transformed and retrieved daily by our data systems, from which data is extracted and repurposed as csv-files.

No such regular national validation process. The act of validation and the extent of it is done at a case by case basis as seen fit by the person extracting the data. 

There are many cases of using several variables in conjecture to group data used on national level.

Not applicable.

No adjustments made.

No adjustments made.