Quality report: Child welfare

Data description

Description of statistic

The child welfare statistics of the Finnish Institute for Health and Welfare (THL) contain information on children and young people who have been placed in out-of-home care or who have been child welfare clients in open care.

The data on children in out-of-home care are based on personally identifiable data. The data on the client relationship and notifications are based on figures. Wellbeing services counties submit the data by municipality on an annual basis. 

The data for the previous year are published annually in April/May. The statistics have been published since 1991. The statistical data are collected on the basis of the THL Act.

Four statistical reports are published of the statistics: out-of-home care, child welfare clients, child welfare notifications and an annual summary report. In addition to the statistical reports, the statistical data are published in THL’s reporting services and on THL’s website.

Relevance

The statistics aim at providing a general overview of the actions taken in child welfare and of the children and young people that child welfare services are targeted at. 

The wellbeing services county-specific and municipality-specific data submitted over a long period of time make it possible to compare different operating practices.

Data content of the statistics

The statistics provide a comprehensive description of client numbers and placements in child welfare in the whole country and in the wellbeing services counties. The statistical data are collected from different sources, which means that not all data are available in all dimensions or as far back in time. The data are proportioned to the size of the population and age groups. In addition, information on issues essentially related to child welfare, such as data on the clients of services for families with children and the costs of child welfare, has been collected to the statistics.

The key concepts and variables used in the statistics have been described as part of the statistical report in the chapter Concepts and definitions.

Statistical process

Source data

THL’s Register of Child Welfare

A client-specific collection of data carried out since 1991. Wellbeing services counties (previously by municipalities) are the producers of the data.

The data in the register is collected on children and young people whose placement has been a support measure in open care, an emergency placement, a placement into care or an after-care measure. The type of the child’s or the young person’s placement location and the duration of the placement are also recorded in the register.

More information about the data content is available on the Instructions on how to submit data page (in Finnish) and on the Data Resources Catalogue page (in Finnish).

Statistical data on open care in child welfare

An aggregated (not client-specific) collection of data obligating wellbeing services counties since 2008. Wellbeing services counties (previously by municipalities) are the producers of the data.

The data are submitted on the basis municipality and mainly by age group. In addition to the number of clients, data is submitted on child welfare notifications, investigations of the need for child welfare, client plans and client relationships that have ended.

More information about the data content is available on the Instructions on how to submit data page (in Finnish) and on the Data Resources Catalogue page (in Finnish).

Instructions on how to submit data (in Finnish)

Data Resources Catalogue (in Finnish)

Other data used for the statistics: 

Data collection method

Wellbeing services counties extract the data from their client information systems and submit the data to THL. The submission of data is described in more detail on the websites described above.

Data validation

The data are checked at THL before publication. Any errors are corrected and data are verified with the data providers as necessary.  THL’s Register of Child Welfare has for decades been dataset that has been widely used in research, monitoring and statistical reporting. Extensive use of the register data has not revealed any systematic problems in the coverage and representativeness of the register. The register has not been subject to any actual validation studies, but involuntary placements are examined in an ongoing study, in which the data is compared to the decisions issued in the judicial system. 

As for the representativeness of the data on open care in child welfare, the data have in the past few years been examined in relation to the data in the datasets Processing times in child welfare and Social worker-to-client ratio in child welfare social work, and the Social welfare monitoring register. These comparisons have revealed inconsistencies, which are discussed in more detail in Chapter 4.3. Consistency and comparability → Comparability of customer numbers in open care. 

Processing of the data

THL’s Register of Child Welfare

Placing wellbeing services county

The data for the Register of Child Welfare should be submitted from the wellbeing services county that is responsible for the child welfare costs (placing wellbeing services county, Child Welfare Act, section 16a). Because the wellbeing services county to the area of which the child is placed (placement wellbeing services county, Child Welfare Act, section 16b) acts as the service enabler, data on the placement are often received from two wellbeing services counties.

Before the health and social services reform, organising was the responsibility of municipalities and the municipality responsible for the costs of child welfare was similarly defined in the Act. Because THL’s Register of Child Welfare has been based on the placing municipalities, the placing municipality principle has not been abolished from the processing of register data between 2023 and 2024. Instead, the data continues to be requested according to which municipality would be responsible for the costs if municipalities were still responsible for organising (=“paying municipality”). However, if the party submitting the data is not able to determine the paying municipality based on previous legislation, it is possible to use the municipality that was the child’s municipality of residence before the start of the placement. Alternatively, the largest municipality of the wellbeing services county in terms of the number of residents can be used as the municipality data.

This means that conflicting and incomplete data concerning the "paying municipality" or placement wellbeing services county are submitted to the register. From these data, THL’s Register of Child Welfare selects only one “paying municipality” for each placement period. The reasoning is carried out in accordance with the following principles:

In the first placements of the child’s life, the placement wellbeing services county is verified separately from the wellbeing services county if the data have been received from several wellbeing services counties.

If the data for later placement periods are submitted from several wellbeing services counties, the paying municipality of previous placement periods is used. If the data for later years are submitted only by the other wellbeing services county, the placement wellbeing services county may change.

If the data are for some other reason separately verified from the wellbeing services county, the “paying municipality” may also be corrected retroactively.    

Age corrections under the Child Welfare Act

In the register, the processing of all emergency placements and cases in which a child has been taken into care is terminated on the 18th birthday. The placement period may continue, but the grounds for the placement should then be “after-care placement” or “open care placement”.

As from 2025, all placements are terminated at the latest with the date of the 23rd birthday. All placements were terminated at the latest with the date of the 25Th birthday between 2020 and 2024. The maximum age for child welfare after-care was already determined as 22 years on 1 January 2024, but the transition rules still made placements of persons aged 23 and 24 possible in 2024. Before 2020, all placements were terminated on the 21st birthday.

Termination of placement periods without an end date

Over the years, wellbeing services counties (previously municipalities) should report an end date to each placement that has once started. Because in practice, the end date is not always received, different processing rules for terminating placements have been drawn up for the Register of Child Welfare since the beginning of the register. 

Each continuing placement must be reported to THL's Register of Child Welfare every year. If the placement is not reported, it is terminated with the date of the last day of the preceding year. Example:

  • In spring 2019, the municipality reported that a specific child was taken into care on 21 April 2018 and that the placement continued at the end of 2018.
  • In spring 2020, the municipality reported that the same placement that began on 21 April 2018, when the child was taken into care, continued at the end of 2019.
  • In spring 2021, the municipality no longer reports the placement data of the child concerned. In this case, the placement that began when the child was taken into care on 21 April 2018 is terminated with the date 31 December 2020 in THL’s Register of Child Welfare. 
  • In spring 2022, the municipality again reports the continuation of the placement that began on 21 April 2018. The end date 31 December 2020 is then deleted from the data, and in THL’s register, the placement is considered to have continued without interruption from 2018, when the child was taken into care.

Placements are also terminated in relation to more recent data submitted by wellbeing services counties. Changes in the client information systems of municipalities and wellbeing services counties make it possible to submit specified data on the end date, placement location or grounds for the placement to THL several years later.

The implementation of the above-mentioned corrections has not always been successful. Each year, changes occur in the processing of data that also retrospectively affect the data from previous years. The most difficult cases are those in which the register receives both information about a completed placement and an ongoing placement. In these situations, factors such as the child’s age, the grounds for the placement, the placement location and the possible impact of the wellbeing services county’s information system are taken into account.

Every year, about 3–7 per cent of sequenced placements, or placement episodes, end on the last day of the year. However, the commonness of the end date is not entirely explained by the corrections made by THL. Instead, it is likely that the last day of the year is also selected as the end date in practice somewhat more often.

Placements are also terminated in accordance with the age limits referred to in the Child Welfare Act, as described above. If necessary, THL also terminates the placement if the person dies.

If a child or a young person has been placed more than once during the same year, the placement information is published according to the latest placement municipality. Also, if the child or young person has been placed on different grounds and/or in more than in one placement location during the year, the placement information is published according to the latest placement.

Corrections to the dates of placement periods

A person can have only one ongoing placement period in THL's Register of Child Welfare at a time. However, in real life, the child may have been taken into care and is in family foster care, for example, but there is an urgent need to place the child in residential care for one month. In such cases, the placement location in the register is changed to residential care for one month, after which foster family care continues.

It is also characteristic of THL’s Register of Child Welfare that, because of different decisions and entries, placements are not completely sequenced. As a result, the previous placement may end on 23 April and the next one begin on 25 April, for example. Such date information is not changed by THL’s register.

Corrections to the grounds for placement

When data on placement periods is conflicting, the data concerning the grounds for the placement are corrected so that if the child has been taken into care, an emergency placement or a placement as a support measure in open care cannot interrupt taking into care. In conflicting data, involuntary taking into care can also not be changed to non-involuntary. However, the above-mentioned corrections are not made in the entire data, but the principles are limited to the processing of the conflicting data instead.

Corrections to the type of placement location

If the wellbeing services county does not submit data on the type of the person's placement location (e.g., placement in residential care), the placement type of the preceding 12 months is entered in the register. If the person has not been in out-of-home care during the previous year and the wellbeing services county is unable to specify the placement location, placement location type “other” is entered or the party who submitted the data is asked to correct the data. Retroactive checking of data with wellbeing services counties will be developed during 2026. 

Statistical data on open care in child welfare

THL uses the submitted data to calculate key figures, such as averages and proportions. Because the data are compiled data, the possibilities for their further processing are limited.

Data revision

Wellbeing services counties often correct the placement data they have submitted in previous years. The number of persons in out-of-home care during the most recent year is typically supplemented with about 100–200 persons in the statistics of the following year, and the data of the previous years are also updated to THL. These corrections also result in changes to previously published annual time series data.

If necessary, municipalities also correct the data on open care retrospectively. 

Quality assessment of the statistics

Accuracy and reliability

The data in the statistics can mainly be considered very reliable. Reliability is based on the following factors: 

The data content has been similar for a long time. 

THL examines the reliability of the submitted data in relation to data that have been submitted before. 

Because the data is used widely, possible errors and irregularities are detected quickly.  

The placement data in the Register of Child Welfare are especially reliable because the data are received with identifiers and possible errors can therefore be traced and corrected with the help of personal identity codes.

Because the data on open care are not received with personal identifiers, they cannot be considered as reliable and accurate as the placement data. Challenges are discussed in more detail in Chapter 4.3 Uniformity and comparability

Timeliness and punctuality

The child welfare statistical data are published annually in April/May. Data is mainly reported on the previous calendar year. In addition to the child welfare statistics, THL publishes statistics on child welfare also regarding processing times in child welfare and the social worker-to-client ratio in child welfare social work every six months. The statistics are developed to be more up-to-date using data from the new Social welfare monitoring register so that the data will be published several times a year. 

Coherence and comparability

Register of Child Welfare

Impact of changes in client information systems

Wellbeing services counties are introducing client information systems that comply with the Client Data Act and cover the entire wellbeing services county between 2023 and 2026. The changes cause conflicting data, from which THL processes unambiguous register information. The more conflicting data THL receives, the more likely it is that some of THL’s processing does not correspond to reality. On the other hand, due to the introduction of the Kanta services and the expansion of the Social welfare monitoring register, THL is also able to validate placement data in new ways.

The most significant single challenge relates to the termination of placements and especially taking into care before the 18th birthday, which are few in number. In connection with changes to client information systems, previous placement periods have been terminated in systems taken out of use, making it difficult for THL’s Register of Child Welfare to determine in which case the information about the end of a placement was genuine and in which case it was due to a change in the client information system.

Comparability of data on client numbers in open care

The information on open care in child welfare is submitted at the level of figures without personal identity codes. The data in the statistics are based on extractions which are made by wellbeing services counties (and previously municipalities) and information system suppliers and include regional differences.

Wellbeing services county-specific comparability: data on open care 

Especially the number of clients in open care is determined in different ways in the wellbeing services counties. In some counties, the data extracted include persons who have during the year received child welfare services that are typical specifically of open care, such as intensified family work or a named support family. In other counties, the client relationship in open care has been understood in a broader sense to include all child welfare social work. According to the latter definition, children who have been placed in out-of-home care or have been taken into care also receive open services. In this case, the number of clients in open care is the same as the number of child welfare clients. 

THL has not specified the rules for extracting open care data care data, as there are significant changes to client information systems between 2022 and 2027. Several changes are underway simultaneously in child welfare data management. Client information systems are being reformed as the recording of client data is transferred to the Kanta services. The comprehensive reform of the Child Welfare Act has also been anticipated to have impacts on statistical reporting.  In this situation, the development of information proceeds in two parallel ways. On the one hand, the aim is to keep current data collections as unchanged as possible so that the time series of wellbeing services counties remain comparable. On the other hand, new information production is being built on the Social welfare monitoring register based on the Kanta services.

Information must be stored in the Kanta services, from where THL will receive the data for the Social welfare monitoring register. In addition, wellbeing services counties have had to submit information containing identifiable data on child welfare clients to the monitoring register since 2023. In the next few years, the quality of the data will therefore be developed especially by beginning to utilise the data in the Kanta service. In addition, potential comprehensive reforms to the Child Welfare Act and its impacts on data management have been waited for. With these changes taking place, data on child welfare is most reliable if the time series do not change.

Although the activities of the wellbeing services counties are governed by the same legislation, the counties apply their obligations in different ways and there are local differences in the service systems. This is further reflected in the comparability of the data. 

Table 3 in the statistical report on child welfare clients shows, by wellbeing services county, an estimate of whether the number of open care clients includes all child welfare clients. The estimate is based on a comparison of the annual open care data and the number of clients on a specific day collected in the data collection on the social worker-to-client ratio in child welfare. If the number of clients on a specific day is clearly higher than the number of open care clients for the whole year, the number of open care clients cannot include all child welfare clients.

Temporal comparability: open care

Apart from individual exceptions, the key data in the collection of data on open care, such as the number of child welfare notifications and the number of open care clients, has been comprehensively received for a long time. The possible missing data have been described annually in the special issues section of the quality descriptions. 

The current Child Welfare Act entered into force on 1 January 2008. Along with the Child Welfare Act, the data content of the national child welfare statistics was also renewed as a whole. The reforms have been described in the quality descriptions of the statistics for the subsequent years. 

The Social Welfare Act that entered into force on 1 April 2015 and changed the definition for a child welfare client has a significant impact on the data on open care. As from 2015, the client relationship in child welfare begins when the assessment of service needs shows a need for child welfare or child welfare services, or urgent measures are taken as a result of a filed child welfare matter to secure the child’s health and development, or when services referred to in the Child Welfare Act or other support is given to the child or the child’s family before the assessment of the service need is completed. Before the change, the client relationship in child welfare began when the investigation of the need for child welfare was started.

In the statistical reports preceding the Child welfare 2021 statistical report, those aged 18 and over were not included in the number of children subject to a child welfare notification. The time series in the tables and indicators of the statistics has been changed retroactively.

Clients turning 18 mainly fall into the age group 18–22-year-olds in the statistics if the client relationship continues and/or the information systems calculate clients’ age at the end of the year. If the client relationship ends and the information system calculates the clients’ age at the end of the client relationship, these young people may in some places be included in the statistical age group of 0–17-year-olds. 

The upper age limit for child welfare after-care was 21 until 31 December 2019, 25 years from 1 January 2020 to 30 June 2024 (including transition periods) and at most 22 years from 1 July 2024.

Comparability of placement data

THL (National Research and Development Centre for Welfare and Health Stakes until 31 December 2008) has collected data to the child welfare register since 1991. The concepts and definitions of the data collection have remained practically the same during the entire history of the register. 

Wellbeing services county-specific comparability: placement data

Legislation defines the placements in child welfare in so much detail that there are not any systematic regional differences in the data in THL’s Register of Child Welfare. However, with regard to some details, such as reporting after-care support as a placement (e.g., supported housing), there may also be regional differences in submitting the data in addition to differences in the actual activities.

Temporal comparability: placement data

Apart from a few exceptions, comprehensive placement data have been obtained for a long time. The possible missing data have been described annually in the special issues section of the quality descriptions. 

The maximum age for child welfare has been changed several times during the 21st century. This is why the reference population of placements of young people is always described in a case-specific manner.

In the statistical reports preceding the Child welfare 2021 report, some of the statistics for 0–17-year-olds did not include those placed in care who turned 18 during that year. Currently, the statistics for 0–17-year-olds contain all of those children who were underaged when placed in out-of-home care during the year. The change has been implemented retroactively to the tables and indicators of the statistics. 

Institutional mandate

The production of the statistics is based on the Act on the Finnish Institute for Health and Welfare (688/2008) and the Act on Statistics (280/2004). One of the official tasks of the Finnish Institute for Health and Welfare is to produce statistical data on the health and welfare of the population, the factors affecting them, and the use and functioning of healthcare and social welfare services to support decision-making, development and research. THL’s statistical production practices are guided by the instructions, recommendations and regulations of Eurostat and the Official Statistics of Finland as well as the principles of statistical ethics.

Data sharing and publishing

Four statistical reports are published of the statistics: out-of-home care, child welfare clients, child welfare notifications and an annual summary report. In addition to the statistical reports, the statistical data are published in THL’s reporting services and on THL’s website.

Confidentiality

As an authority, the Finnish Institute for Health and Welfare is tasked with reporting collected data regarding national health and welfare. The data used to draw up THL’s statistics is primarily confidential, and personal data may not be published. The protection of processed data is based on the Act on the National Institute for Health and Welfare (688/2008), the Statistics Act (280/2004), the Act on the Openness of Government Activities (621/1999), the EU General Data Protection Regulation (EU) 2016/679 and the Data Protection Act (1050/2018) as well as other regulations guiding the activities of the Institute.

THL’s datasets are secured at all stages of processing. Data and information systems can only be accessed by persons who have a permit to access certain data for clearly defined purposes. Others do not have the ability to view, process, change or delete data. Written instructions have been drawn up for ensuring the data protection of completed statistics. All THL personnel are bound by a confidentiality obligation.

Special issues in the 2025 statistical reports

Out-of-home care 2025

The 2025 data were received from all wellbeing services counties, for residents of all municipalities.

Changes in client information systems had an impact on the quality and comparability of the data:

  • In North Ostrobothnia, in municipalities using the Saga system, there were significant shortcomings in the termination of placement periods. THL has automatically terminated all emergency placement periods 30 days from their start. Missing end dates may result in more placement days being recorded in the child welfare statistics than there actually were, and more children being recorded as placed on 31 December than there actually were. North Ostrobothnia will verify the 2025 placement data later in 2026.
     
  • In Southwest Finland, the transition to a single information system has taken place. In this context, previous placement periods have been terminated in systems taken out of use with the date 18 December 2025. In THL’s Register of Child Welfare, it has been difficult to determine in which case the information about the end of a placement was genuine, but the end date information has mainly been entered in THL’s register. Potentially excessive end dates may result in fewer children being recorded as placed on 31 December than there actually were. Because the end dates are close to the end of the year, the impact on the number of placement days is minor. Excessive end dates may also have an impact on national data on taking into care that ended before the 18th birthday, which are few in number.
     
  • In some areas, the differentiation of small municipalities was not possible, so the data for two municipalities were reported under one home municipality code.

Child welfare clients 2025

The 2025 data were received from all wellbeing services counties, for residents of all municipalities.

In some areas, the differentiation of small municipalities was not possible, so the data for two municipalities were reported under one home municipality code.

  • South Karelia: all data under Lappeenranta.
  • Pirkanmaa: data for Kuhmoinen under Kangasala.
  • Utsjoki and Inari were reported as a combined total under the Inari municipality code.

Changes in client information systems had an impact on the quality and comparability of the data:

  • In Southwest Finland, the transition to a single information system has taken place. Due to the combination of data from different information systems, there are significant uncertainties related to the reliability of the data.
  • In North Ostrobothnia, two information systems are still in use, which means that data on the number of new clients and after-care clients, and the number of client plans and service needs assessments, are not reliable.
  • In Satakunta, a transition to a unified information system took place and the data are not fully comparable with data from previous years.

Child welfare notifications 2025

The 2025 data were received from all wellbeing services counties, for residents of all municipalities.

In some areas, the differentiation of small municipalities was not possible, so the data for two municipalities were reported under one home municipality code.

  • South Karelia: all data under Lappeenranta.
  • Pirkanmaa: data for Kuhmoinen under Kangasala.
  • Utsjoki and Inari were reported as a combined total under the Inari municipality code.

Changes in client information systems had an impact on the quality and comparability of the data:

  • In Southwest Finland, the transition to a single information system has taken place. Due to the combination of data from different information systems, there are significant uncertainties related to the reliability of the data.
  • In North Ostrobothnia, two information systems are still in use, which means that data on the number of new clients and after-care clients, and the number of client plans and service needs assessments, are not reliable.
  • In Satakunta, a transition to a unified information system took place and the data are not fully comparable with data from previous years.

Updated: