The data is collected continuously in the months before the publication from the public budget systems and other supplementary sources. It is then compiled according to national accounts principles, where it may be necessary to contact the specific source to clarify certain characteristics of the entries. It may be necessary to make imputations in cases where the data isn’t available at the time of publication. When a full dataset is compiled for all subsectors balancing is carried out to secure internal consistency.
March-version: Accounts from the previous year and the appropriation accounts for the following year, provisional accounts for local governments for the provisional year and budgets for the following year. For both years budgets is used for LG (the Employees' Guarantee Fund) and enumeration of the latest accounts is used for the unemployment insurance funds (relevant transactions are balanced with information in the appropriation accounts for the state). The figures on extra budgetary units are enumerated on the basis of earlier accounts and information on transfers in the appropriation accounts.
August-version: The Budget, which sets out estimated budgets of local governments, budget figures of LG (the Employees' Guarantee Fund) and enumeration of the latest accounts of unemployment insurance funds (relevant transactions are balanced with information in the appropriation accounts for the state). The figures on public quasi corporations are enumerated on the basis of earlier accounts and information on transfers in the national budget.
Data is received electronically from the central government, the municipalities and the regions' financial management systems.
The received data is from the official budget that have been intensively validated.
When the data is collected and loaded into the database the classification according to ESA2010 starts. This means that expenditure and revenue is classified into categories like e.g. salary, investments, income transfers and interest. The classification is carried out in three steps. In the first step the actual data is merged with datasets for previous years and all classifications on identical account numbers is copied to the new dataset. in the next step new entries on accounts which have a straight forward ESA2010 interpretation classified automatically. The remaining entries are classified manually. Furthermore all entries are classified by COFOG and branch. The COFOG classification is carried out according to the international COFOG manual (Classification of the Functions of Government). This classification classifies into categories like e.g. defense, health and education.
Primary data is classified on the most detailed level to obtain a nearly perfect link between all entries in the primary accounts and the government finance statistics.
When the classification process is finished the compilation starts. The compilation covers various calculations, imputations and time adjustments. This compilation is necessary for the conversion from primary accounts into national accounts. The calculation of use of fixed capital is an example of a calculation which is carried out for the construction of the government final consumption expenditure. Imputations are made for civil servants salaries to reflect the obtained right to pensions in the future. Time adjustments are heavily used in relation to tax revenue due to the fact that according to ESA2010 the tax revenue should be on an accrual basis.
As a final step all transactions between subsectors are balanced and consolidated.
There is no correction of data beyond what has already been described under data validation and data treatment.