Efficiently Retropolated and Quality-Assured Supply-Use Tables

forthcoming

This project aims to develop and test an efficient, automated method for retropolating Supply and Use Tables (SUTs) following national accounts benchmark revisions, with built-in support for expert oversight and quality assurance. It directly supports the objectives of Project 11, subarea (d) of the SMP-ESS-2025-NA-BOP-IBA call. Currently, the rebalancing of historical SUTs in Statistics Denmark is a manual, resource-intensive process, which limits both the number of years that can be revised and the ability to validate results in depth. With the 2029 benchmark revision on the horizon, a more efficient and transparent approach is required. The project will deliver a modular Python-based tool built on GRAS-based rebalancing algorithms. It will include user-defined controls such as fixed cell values and diagnostic features to support plausibility checks. A key innovation is the integration of structural information from a manually rebalanced reference year, allowing that structure to guide the automated rebalancing of earlier years. The tool will be tested on two datasets: • 2014–2020 SUTs, already revised during the 2024 benchmark, to allow comparison with manual results; • 2010–2013 SUTs, which have not yet been revised, to simulate real-world use in a future benchmark cycle. Results will be summarized in a final technical report. If the quality is satisfactory, the revised 2010–2013 SUTs will be transmitted to Eurostat. The tool will be made openly available – preferably via GitHub – provided it is sufficiently generalizable. If not, access will be granted upon request.