Overview of GDP flash estimation methods : 2016 edition
European GDP flash estimates, coordinated by Eurostat, are currently released at t+45 days. To investigate whether the deadline for release could be cut to t+30 days, and to share experiences among Member States, Eurostat set up a task force in May 2013. Sixteen EU countries and Switzerland took part in its work in 2013, 2014 and 2015. There was a particular focus on the methods used to estimate GDP, and this document is the main outcome of a dedicated sub-group of the task force. It gives a general overview of estimation methods used in different EU countries, and presents a number of practical suggestions. After the introductory chapter, the document discusses the preliminary analysis of available data, model strategies, and methods for forecasting missing data, including extensions to multivariate setups and the analysis of results. Two main estimation strategies have been identified: ⎯ the 'direct' approach: the last quarter of a given GDP component is estimated directly, and ⎯ the 'indirect' method: the last quarter is extrapolated using a temporal disaggregation procedure. The difference is given by the model used for extrapolation. Under the former approach, the set of related data is directly used to extrapolate the last quarter of the GDP component. Under the latter, extrapolation is carried out in two steps: firstly, the indicator in the last quarter is extrapolated, then the temporal disaggregation for the GDP component of interest is run. Following the same logic, when no indicator is available a one- or twostep time series method is applied, for the direct or the indirect approach respectively. A summary of the answers to a questionnaire on estimation methods is also presented in the annex to this document. The questionnaire helped researchers understand Member States' experiences, their approach in terms of accounting method, detail of compilation, price evaluation, type of adjustment on data, whether a preliminary analysis is effectively carried out, details of the more commonly used indicators and, of course, the more technical aspects of the methods used. In short, this document is an attempt to merge differing experiences in different EU countries into a unified view. It allows the reader to follow a step-by-step approach to flash GDP estimation, with explanations and analysis to help them conduct the exercise efficiently. The approach is pragmatic, with links to available software, references to the literature, and a particular stress on the elements shared by advanced users and beginners. The aim is to share information about the technical aspects of estimation methods within the community of people working in the area of quarterly accounts. The best practices presented here are thus a way of standardising flash GDP estimation at national level and, indirectly, of improving the quality of European aggregates released after a short period of time.
Year of publication: |
2016 ; 2016 edition
|
---|---|
Institutions: | European Commission / Statistical Office of the European Union (issuing body) |
Publisher: |
Luxembourg : Publications Office |
Subject: | Nationaleinkommen | National income | Statistische Methode | Statistical method | Schätztheorie | Estimation theory | Schätzung | Estimation |
Saved in:
Saved in favorites
Similar items by subject
-
The likelihood test under non-standard conditions : testing the Markov trend model of GNP
Hansen, Bruce E., (1991)
-
Hansen, Bruce E., (1995)
-
Radici, unitarie e persistenza : l'analisi univariata delle fluttuazioni economiche
Pistoresi, Barbara, (1997)
- More ...
Similar items by person