Quarterly national accounts statistics are crucial for shaping economic policies and often require rapid publication of data. At many statistical offices, a significant portion of the calculations is still performed at least partially manually, which is time-consuming and prone to errors. This master's thesis therefore focuses on the automation of the quarterly gross domestic product (GDP) calculation process at the Statistical Office of the Republic of Slovenia. The aim was to develop an automated solution that would enable faster, more accurate, and more transparent preparation of GDP estimates. The automation process included the design of a central database using Oracle SQL Developer, the development of ETL procedures in SQL Server Integration Services, the use of the R programming language for statistical estimation, and the development of calculation processes in SAS. The thesis focused on the production approach to GDP calculation, addressing the challenges of integrating various administrative and statistical data sources. The results show that automation reduces processing time, minimizes the risk of errors, and improves traceability and consistency of data, although some steps still require manual review. The thesis confirms that automation significantly contributes to the modernization of statistical processes and provides a foundation for further enhancements towards greater methodological flexibility and faster responsiveness to changes in data sources.
|
