Statistics Finland

Economic Statistics

Statistics Finland, Economic statistics, Työpajankatu 13, FI-00022 Statistics Finland, Finland

The quarterly business demography measures the number of newly registered legal units and the number of legal units that have started the process of declaring bankruptcy. The statistics are based on existing administrative data.

Statistical classification of economic activities NACE Rev. 2.0

The data covers all legal units in market activities listed in NACE sections B to N, P to R and Divisions S95 and S96.

The variables are measured in absolute values and they are based on administrative data.

Legal unit

The data covers all legal units in market activities listed in NACE sections B to N, P to R and Divisions S95 and S96 that have been registered or started the process of being declared bankruptcy during the reference period. 

The data covers all national legal forms that can be thought to provide services or goods to the market. 

Finland

From 1^st Quarter 2020 on

Not applicable

Absolute values

Quarter

Summary

The Council Regulations (EC) concerning short-term statistics The Statistics Act (280/2004, amendment 361/2013) The Statistics Act (280/04) is the general act for the National Statistical Service (NSS). It contains the principles for the data collection, processing, and dissemination of official statistics. The act defines the roles of statistical authorities (Statistics Finland, Customs, National Institute for Health and Welfare, and Tike, Information Centre of the Ministry of Agriculture and Forestry) and other authorities producing statistics. The Statistics Finland Act (48/1992) states that Statistics Finland (SF) shall provide for the general development of official statistics in collaboration with other central government authorities.

The aim of the NSS is to produce official statistics, Official Statistics of Finland (OSF). European law (especially the Regulation of the European Parliament and of the Council on European statistics (EC) No 223/2009) applies to a large portion of OSF. Statistics Act (280/2004) | Statistics Finland, http://tilastokeskus.fi/meta/svt/index_en.html 

European legal basis: REGULATION (EU) 2019/2152 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 27 November 2019 on European business statistics, repealing 10 legal acts in the field of business statistics (CELEX 320R2152), Table 1 of the Commission implementing regulation 2020/1197.

Ensuring statistical reporting

The Statistics Act requires that the primarily exploited sources for statistical purposes shall be the data accumulated in the administration of general government and the data produced as a consequence of the normal activities of employers, self-employed persons, corporations, and foundations. SF has the right to have access to these data under the Statistics Act. In addition, all public and private entities in Finland are obliged to provide SF with data on their finances, products and staff as necessary for the production of statistics. The right of SF to collect data by virtue of the obligation does not extend, however, to data that are kept confidential for reasons of international relationships, public safety, the interest of national defence, or the safety of the state. Before any data collection based on the obligation, the statistical authority must consult the respondents or their representatives about the contents, timing, collection methods, duration of storing of the data as well as about costs. Besides those data obtained directly from administrative sources data from natural persons are always collected on voluntary basis by using interviewing or mail and web surveys. In addition, the interviewees must be informed in advance in a written form. 

The Statistics Act stipulates that a data provider who willfully fails to provide the obligatory data or willfully provides false data shall be sentenced to a fine. Nevertheless, SF is allowed to refrain from bringing charges if the violation is regarded as minor, but in practice charges have not been filed.

Data sharing and coordination among data producing agencies

According to the Statistics Act, data obtained by four statistical authorities may be released to other parties either if permitted by legal provisions explicitly concerning the NSS, or upon express consent of the subject of the data. As far as statistical authorities are concerned they are allowed to transmit confidential data with identifiers to each other if it is deemed necessary for the production of statistics. The same applies to the European Statistical System authorities (ESS Authorities). Co-ordination among data-producing agencies is normal practice at both specialist and top level. 

Absolute values of bankruptcies and registrations are reported quarterly to Eurostat. 

Confidentiality of individual reporters' data

Several legal acts guarantee that individual data should be kept confidential.

According to the Act on the Openness of Government Activities (621/1999) data collected for statistical purposes are confidential. The rule is not applied to the publicity of data describing the activities of central and local government authorities and production of public services or to certain data in the Register of Enterprises and Establishments.

Under the Statistics Act, statistics shall be compiled so that the respondents whom they concern are not directly or indirectly identifiable. Violation of the secrecy obligation is punishable under the Penal Code. At the EU level, similar assurances are included in Council Regulation (EC) No 223/2009.

Details regarding the protection of information on private individuals are laid down in the Finnish Personal Data Act (523/1999).

The Statistics Act obliges statistics-producing authorities to inform respondents in writing or in other appropriate manner about the intended use of the data, the procedures to be used in producing the statistics, the principles governing whether the provision of data is obligatory or voluntary, the rights of the respondents, the arrangements for protecting the data, and the duration the data will be stored.

The Statistics Act allows a statistical authority to grant access to confidential data for use in scientific research or statistical surveys if statistical units cannot be identified directly from them. The right to use data may be given in compliance with a well-defined process including a written application.

Statistics Finland (SF) has implemented procedures to prevent disclosure of any individual data provider. It has published guidelines on how to apply the Statistics Act and the Personal Data Act, as well as guidelines on the protection of tabulated data on enterprises and individual persons. A section on data protection is included in the SF publication Quality Guidelines for Official Statistics. Micro data concerning individual persons released for scientific research are first edited to remove variables that would make it possible to directly identify individual persons such as name, address or personal identification number. Similar procedures are used with respect to sensitive information about units other than individual persons.

https://tilastokeskus.fi/meta/tietosuoja/index_en.html

Confidentiality and reference to the Statistics Act are announced in the data collection questionnaire. Links to different acts on statistics and ethical principles are presented on the website of Statistics Finland. The guidelines on professional ethics of Statistics Finland have been published (Handbooks No 30, Statistics Finland, 2002).

Not applicable

Release calender can be found here: https://stat.fi/en/publications

The data is not nationally disseminated in the same form as the data sent to Eurostat. Bankruptcies are part of the statistics: "Bankruptcies and business restructuring proceedings" while registerations are part of the statistics: "Enterprise openings and closures".

https://stat.fi/en/publications

The data is not nationally disseminated in the same form as the data sent to Eurostat. The data released follows general guidelines of the Statistics Finland. Release dates can be found from release calender and data is disseminated for all users at the same time.

The data is not nationally disseminated in the same form as the data sent to Eurostat.

The data is transmitted to Eurostat quarterly.

Not applicable

Not applicable

Not applicable

Access to microdata can be requested through Statistics Finland research services following guidelines found here: https://www.stat.fi/tup/mikroaineistot/index_en.html

The data are sent to Eurostat

Not applicable

Not applicable

Dissemination of documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques

The quality description of the statistics, as well as concepts and definitions (in Finnish, English and Swedish), are available on the home page of the statistics. The quality description of the statistics is included in every publication.

While compiling statistics, Statistics Finland observes the European Statistics Code of Practice (CoP).

Impartiality of statistics

Section 11 of the Finnish Statistics Act contains the requirement that statistics should be as reliable as possible, give a truthful picture, and make use, if possible, of uniform concepts, definitions, and classifications, as well as be timely. Similar provisions can be found in EU legislation. The Statistics Act, the Statistics Finland Act, and the Council of State Decree on Statistics Finland (1063/2002) and several other legal provisions support SF’s professional independence:

Section 3 of the Statistics Finland Act decrees that the internal organisation of SF is defined in internal Rules of Procedure which are decided by SF’s Director General (DG);

Section 1 of the Council of State Decree on SF gives the DG the right to manage the activities and finances of SF;

Section 5 of the Council of State Decree lays down that the DG shall be appointed by the Council of State for a fixed term. The nomination is made after open competition. The Decree also gives the DG the right to appoint the staff of SF, including directors of the departments as well as other staff members unless the decision-power has been delegated in the Rules of Procedure to some other official of SF;

Section 10 of the Statistics Act provides that data processing shall take place in accordance with good statistical practice and the international recommendations and procedures generally applied in the field of statistics.

Selection of sources, methodology, and modes of dissemination

The selection of sources, methodology and modes of dissemination of SF statistics are decided by SF alone. According to the Statistics Act, the choice of data sources, statistical techniques, and methods of dissemination should be based on statistical considerations (Sections 3, 10, 11, and 12). There are also responsibilities regarding cost-efficiency and the response burden of data collections (Section 4).

Guidelines for staff behavior

On the basis of international recommendations SF has confirmed its own Guidelines on Professional Ethics. All employees are given a copy of this handbook. An internal ethical board meets regurlarly and reports on its decisions and discussions on the intranet open to all staff of SF.

Quality monitoring

The top management of SF has made several self-assessments in line with the EFQM model. There have also been external audits by e.g. the EU and IMF experts. Processes are in place to monitor the quality of the statistical process and the processes of individual statistics. Quality considerations are an integral part of the planning and evaluation of the statistical program. The process owner of statistical production and it’s supporting group monitor the quality and steer the standardisation of work processes. Statistics Finland has an internal quality audit system. The main objectives are to evaluate the ways of working, methods and techniques. An audit is carried out by an audit team of experts who are external in the sense that they do not have any direct connection with the production process in question. About 8 audits are carried out yearly.

Main users:

Government
Research institutions
Eurostat
ECB
IMF
Market analysts
Enterprises
Media
Main uses:
Economic follow-up and forecasting
Analysing business cycles

Monitoring user requirements

Co-operation between SF and important users with regard to the relevance of statistics and the users’ needs consists of an extensive feedback system and co-operative working groups with the main users, such as users of national accounts. There are regular meetings of SF directors and experts with the users, even at the senior management level. Users are usually also invited to participate in discussions concerning the establishment of new statistics or revisions of existing ones. In addition, there are specific feedback systems for receiving the users’ opinions at SF. These systems consist of an anonymous feedback channel on the web, media monitoring, surveys among different user groups for the evaluation SF’s performance, user surveys (every second year), and a system for collecting and disseminating information that is strategically important for SF. Specific statistical products conduct their own user surveys and keep in regular contact with their main interest groups.

All the required series are produced.

In the registration process, the founder receives the business identity code from PRH, after which the founder notifies the operating industry to the Finnish Tax Administration. The Finnish Tax Administration then updates the industry code to the database from which we will compile the registration data. The lag between the registration (receiving the Business ID) and the updating of the industry code is the reason why, at T + 40, there is a significant number of newly registered legal units without industry codes. The data is corrected using verbal description of the activities and a classification algorithm. The newest quarter will most likely revise but the methodological correction makes the revisions smaller and the revisions are monitored constantly.

Not Applicable, since data collection is a census.

Not Applicable

The data is transmitted to Eurostat at latest 40 days after the reference period.

No delays have occured

Data are comparable between countries. Results are based on the requirements of the EU Regulation, which is in use in all EU Member States. 

Due to covid-19, changes into bankruptcy legislation were made in 2020Q2 and 2021Q1 whic has affected the number of bankruptcies. in 2020Q2 the legislation was temporarily changed so that it was more difficult for claimants to file a petition to delare legal unit bankrupt. Therefore, the number of bankruptcies decreased significantly. The legislation was changed back to normal in 2021Q1.

Not applicable

Not applicable

There are no burden on respondents. The cost of the production consists of only personnel costs.

The newest quarters of registrations will be revised due to the lag and estimation of NACE codes. The same revision policy is applied nationally and in transmission to Eurostat.

The revisions will be made in the data transmission schedule. Typically small revisions will occur and in case of major revisions data would be manually checked and Eurostat would be informed.

The source data for registrations is Patent and Registration Office (PRH) data on registrations of new legal units. The source for bankruptcies is Legal Register Centre (ork). The data is supplemented with information from the national business register.

Continuous

PRH transmits the data daily to Statistics Finland and ORK transmits data monthly. The quality of the data transmission is monitored continuously by analysing the number of legal units in the data.

The number of legal units in the statistics is monitored closely. Additionally, we calculate percentage changes from previous periods and see where the changes occure and if they are reliable. Manual checks are also made to the bankruptcy data.

The data is compiled by collecting all legal units that belong to the statistics. The number of these legal units is then aggregated to the NACE levels required.

To address the problem of missing NACE codes, we implemented a classification algorithm to the data. We created a data set that contained the individual business id, the corresponding verbal description of the activities of the enterprise and the related NACE section marked as X for those legal units that have no NACE class. To obtain NACE sections to the legal units with value X, we built a following process utilizing Natural Language Processing (NLP) and Support Vector Machine (SVM):

1. Cleaning and processing of the text data
In this step the text data is processed so that its predictive power is as high as possible. This means removing of specific words such as stop words, or words that are included in nearly all of the descriptions. Additionally, the text was lemmatized using Finnish dictionary, and only substantives, adjectives and verbs were kept. Also, all punctuations were removed, and all letters made lower case.

2. Document term matrix
After the pre-processing, the data was transformed into a document term matrix (DTM). In DTM a row represents one observation corresponding to one legal unit. Each word in the whole dataset is one column, and the values of the cells are counts of how many times each word appears in each row.

3. Splitting the data
The data was split into three sets. First set contained all the legal units with no original NACE codes, so they needed to be predicted. The remaining legal units for which we had NACE codes, were split into training and test sets.

4. Model fitting
The linear SVM model was trained using the training set and the predictive capabilities of the model were tested using the test set. The overall accuracy of the model is around 55 per cent. 

5. Predictions
Finally, the model was used to predict a NACE section for all the legal units for which we had no NACE code. The predictions are used only for the newest quarter.

6. Correction with confusion matrix

Predictions are corrected using user's confusion matrix. That way we can eliminate systematic errors thus increasing overall accuracy over 80 per cent.

Not applicable

-