BACK TO SCHEDULE REGULAR

Session 24

Improving Comparability and Coherence

4 June 2026
14:15 – 15:45
ŠIBENIK V

Presentation title
Implementation of a methodological Eurostat advice regarding the sector classification of local public transport in Germany
The Maastricht Treaty established a framework for multilateral fiscal surveillance within the European Union.

Read more Read less

This framework is intended to guarantee the financial stability of the member states by providing guidelines on acceptable levels of government spending and debt. Eurostat collects relevant data on government spending biannually, which is then examined by the European Commission. Should a member state deviate from the guidelines, an excessive deficit procedure will be implemented against it.

A key requirement for a meaningful examination is the comparability of data on government spending and debt across member states. This is a challenging task due to the significant heterogeneity across countries within Europe. The responsibility for ensuring the comparability of the data and the appropriate application of common methodological rules lies with Eurostat, which can provide Member states guidance via a so-called bilateral advice.

In 2022, DESTATIS received such a bilateral advice concerning the treatment of short-distance public passenger transport. The advice requested that DESTATIS re-evaluates the recording of payments by government to fund local public transport. Following the re-evaluation, many public transport and rail infrastructure units were reclassified into the government sector. This reclassification was not only prospective but also retrospective, requiring adjustments to the entire time series which entailed methodological and data-related challenges.

This presentation outlines the initial situation prior to the integration, describes the methodological and data-related obstacles encountered, and presents the outcomes of the reclassification process. By addressing these issues, the presentation contributes to the ongoing effort to enhance the comparability and reliability of government spending data within the EU’s multilateral surveillance framework.

Main author / Presenter
Niels Gillmann
Destatis

Read more Read less

Niels Gillmann studied economics in Halle (Saale), Germany, Saint-Denis (Réunion), France and Uppsala, Sweden. He completed his doctorate in time series econometrics at TU Dresden, Germany. Since 2024, he has been employed in the national accounts unit of the Federal Statistical Office of Germany. His responsibilities include local public transport and questions of sector delimitation. Previously, he worked as an economist at the ifo institute in Germany, where he worked on economic forecasting and political advisory services.

Presentation title
Capturing the digital transition in External Statistics: quality challenges in cloud computing and digital intermediation platforms
The widespread digitalization and integration of new technologies are reshaping global work processes as well as cross border commercial transactions, with cloud computing and digital intermediation platforms (DIPs) emerging as pivotal new economic activities with far-reaching implications for measurement and quality assurance.

Read more Read less

This rapid development deeply challenges the quality, consistency, and interpretation of external sector macroeconomic data, impacting traditional economic measurement concepts and practices.

These digital transactions were insufficiently covered in the Balance of Payments and International Investment Position Manual, Sixth Edition (BPM6) and the absence of specific guidance for these new technologies has led to conceptual uncertainties and inconsistent practices across countries. The necessary clarifications are addressed now by the new IMF’s Integrated Balance of Payments and International Investment Position Manual, Seventh Edition (BPM7).

Cloud computing services consist of computing, data storage, software, and related ICT resources delivered on-demand over a network with metered usage, shifting the location of computing from the user's premises to remote data centers.

On the other hand, Digital Intermediation Platforms are online platforms that supply goods or services facilitating interactions between two or more distinct but interdependent sets of users (buyers and sellers) for a fee or commission. DIPs are distinguished from e-tailers by the fact that they do not take economic ownership of the goods or render the services being intermediated. Their multi-sided nature and reliance on implicit charging models further complicate statistical recording and pose challenges for transparency, accuracy, and comparability.

This paper specifically examines the challenges that cloud computing and DIPs activities pose for compiling and interpreting Balance of Payments (BOP) statistics and how potential misallocations of relevant counterpart countries, misclassifications among service categories or not addressing these transactions may result in quality issues in BOP, affecting in a negative way users and policy decision makers needs by providing misleading information.

Acknowledging the complexities involved in accurately capturing the economic activities of cloud computing and DIPs, in August 2025, Eurostat conducted a comprehensive survey among its Member States to gather insights into the current practices, available data sources, challenges, and opportunities encountered by national central banks and national statistical institutes in identifying these platforms and measuring their economic activity. The document will summarize the main outcomes of these surveys and further considerations from Eurostat how some of the challenges can be addressed and overcome to achieve high quality external statistics data.

Main author / Presenter
Viviana Vitali
Eurostat, European Commission

Read more Read less

Viviana Vitali is a Statistical Officer at Eurostat within the National Accounts Methodology unit. She holds a Master’s degree in Data Science and has a background in Mathematics, with over five years of professional experience in macroeconomic statistics, specialising in Sector Accounts and the Balance of Payments. Her current work focuses on the development, interpretation, and EU-wide implementation of the Integrated Balance of Payments and International Investment Position Manual, Seventh Edition (BPM7). She contributes to the production and quality assurance of balance of payments statistics, as well as the methodological classification of international transfers related to the EU Multiannual Financial Framework. She is an active member of the Balance of Payments Working Group and regularly contributes to methodological coordination and discussions within the balance of payments domain. Her research interests lie at the intersection of globalisation and the digital economy, particularly the impact of Digital Intermediation Platforms and cloud computing.

CO-AUTHORS:

Presentation title
An Early Warning System for asymmetries in International Trade in Services Statistics.
Countries have often expressed the need, during Working Group meetings, for the development of an ex ante report on bilateral asymmetries in International Trade in Services Statistics (ITSS), prior to data transmission, in order to better deal with asymmetries.

Read more Read less

The need for such a report comes directly from the practical aspect that countries can perform revisions on ongoing datafiles more easily before dissemination (i.e. during production), than after dissemination, where all accounts have been closed. To fill this gap, Eurostat is currently developing an Early Warning System (EWS) for the annual ITSS dataset. Given the limited resources and the time constraints, an efficient EWS should highlight for each country the most significant asymmetries, as early as possible, and with the highest probability to appear in future submitted files. Annualized (sum of quarters) trade in services Quarterly Balance of Payments (QBOP) data come several months earlier, but at the cost of containing less information (as the mandatory geo breakdown and EBOPS items of QBOP data for services are less than the respective ones in the annual ITSS dataset), more confidential cells and being subject to more frequent revisions. By using historical first file submissions of QBOP and annual ITSS data, we construct a simple statistical measure to capture the probability that “an asymmetry case in the annualized QBOP will actually appear in the annual ITSS data”. This measure allows us to remove resource-wasting asymmetry cases, where countries engage early in an asymmetry case based on QBOP data that in the end would not exist in the annual ITSS data whatsoever. Finally, we combine this probability measure with the existing scoreboard methodology employed in the ITSS Asymmetry Resolution Mechanism (ITSS-ARM), to build an EWS that selects asymmetry cases in an optimal manner. The goal of the novel EWS in annual ITSS is to highlight to countries five (5) months ahead the most important and most probable asymmetry cases, which they should try to resolve, during the roduction phase. This should increase the data quality of the annual ITSS data and reduce future revisions.

Main author / Presenter
Marios Papaspyrou
Eurostat

Read more Read less

Marios Papaspyrou studied Mathematics at the University of Athens and Economics at University of Piraeus and Athens University of Economics and Business. He holds a PhD in Economics, with a focus on Bayesian Econometrics and applications on technical efficiency. For the last 4 years, he has been a statistical officer at Eurostat, in the Unit of Globalization and International Trade in Services, working mainly on resolving bilateral asymmetries and running the International Trade in Services Statistics Asymmetry Resolution Mechanism (ITSS-ARM). Previously to Eurostat, he worked for the National Central Bank of Greece for 12 years at the balance of payments data collection division.

Presentation title
A Model‑Based Framework for Benchmarking the External Consistency of Structurally Lagging Economic Indicators
Ensuring the external consistency of economic time series is essential for national statistical institutes (NSIs) operating under recurrent production cycles.

Read more Read less

To support this process, we are developing an internal quality assurance system for benchmarking key Labour Force Survey (LFS) aggregates—employment, unemployment, and hours worked—against selected external reference indicators. The comparison set includes a wide range of time series from economic sentiment surveys, national accounts statistics and public employment service statistics. A guiding principle is to exploit cycle‑sensitive leading indicators as well as indicators that are conceptually closely related to LFS variables. This combination enables both broad cyclical benchmarking and more direct consistency checks.

This work presents a model-based approach for monitoring short-term changes in key aggregated labour market time series. A Vector Error Correction Model (VECM) is trained with an expanding window to continuously update expectations as new information becomes available. By comparing one-step-ahead forecasts with observed outcomes, the method aims to highlight sudden deviations as well as gradual structural shifts in the data. Two complementary alarm mechanisms are introduced: a pointwise alarm for abrupt deviations and a rolling alarm capturing persistent divergence over time. The approach is designed to support early detection of change and data issues, complementing existing outlier detection methods rather than replacing them.

The system generates fully automated and reproducible reports that are routinely used during the monthly review of preliminary LFS estimates. By providing standardized alarms and transparent diagnostics, the framework helps analysts distinguish genuine labour market developments from potential data irregularities or temporary disturbances. For other NSIs, this approach demonstrates how model‑based monitoring can be combined with leading indicators to enhance internal quality control without altering established production workflows. The framework is extensible to additional indicators and diagnostics, offering a flexible foundation for strengthening the robustness and interpretability of short‑term labour market or other economic statistics.

Main author / Presenter
Stefan Andersson
Statistics Sweden

Read more Read less

The authors work at Statistics Sweden with the Swedish Labour Force Survey (LFS). Petter has a background as a methodologist working with methodological questions concerning the Swedish LFS and Stefan has a background as an economist working with data analysis related to the Swedish LFS.

CO-AUTHOR: