This report on the validation, quality assurance, and demonstration of the new MRIO-tables of Statistics Sweden was produced as part of Swedish development work to use interregional IO-tables for research, policy assessment, and planning. The work constitutes an increase in the level of ambition regarding the production of knowledge support concerning linkages among Swedish regional economies, and internationally, through trade in goods and services.

Questions on what policies will be effective to further develop Sweden’s competitive edge, and to promote Swedish efforts to stall climate change in an ever more specialized international economy are at the centre of the report. It was for long considered infeasible to address these questions with the help of interregional input-output analysis. The report shows that such an increase in the level of ambition is both motivated and possible especially in Sweden with its well-developed economic statistics, and its acceptance that public policy should be evaluated using scientifically sound, and evidence-based methods.

A main result of the report is that there is reason to continue and further develop the development work on interregional input-out tables presented by Statistics Sweden. For this effort to be successful, it is necessary to add further resources for the collection of statistics, the development of quality-assured statistical products, and the demonstration of the strength and weaknesses of these methods in relation to current practice.

A fundamental problem identified in the ongoing MRIO-project within Statistics Sweden, further developed in the research report, is that the regional level is subordinate to the national level as concerns the collection of statistics, and the establishment of economic accounts. National accounts are based on micro data without utilizing the geographic information contained in micro data. After that, the regional accounts are set up using top-down methods of various kinds.

It is shown in the report that this may lead to consistency problems for both internal deliveries in the production system, and for final demand. The report takes private consumption as an example. The treatment of trade and transport margins is given as another example. The report argues that Statistics Canada has better integrated the regional and national levels in particular regarding interprovincial and international trade.

The report points out that additional synergies can be achieved through a further cooperation between ongoing development work in the Government Agencies Growth Analysis and the Swedish Growth Agency as well as within Traffic Analysis and the Swedish National Transport Authority. These can consist of a coordination of collection of primary data for commodity trade and exports. A main result stressed in the report is that these integration issues can be developed further in the third phase of the MRIO project. This is more important than proceeding with add-on evaluation models for areas as competition policy, employment change, and climate policies. Such models already exist, for instance, in the so-called Raps system for regional economic evaluations.

It is also shown that the next phase of the Statistics Sweden project will benefit from focusing on linking the Swedish regional and national accounts with the international work on consolidated input-output analysis for groups of countries. It is noted that thebSwedish accounts earlier have been subject to adjustments, for instance, within the OECD for them to work in analyzing global value chains, and global climate policies. Now adjustments have been made and it is within reach to make Sweden a role model when it comes to evaluating international and interregional policies within the same consolidated framework.

The report was written in close cooperation with the ongoing MRIO-project within
Statistics Sweden. This should guarantee a smooth implementation of the results in the
third phase of the MRIO-project.

The current phase of the MRIO-project has not been finished yet, even if the bulk of the
development work was completed. A major achievement for the future is the coding of
estimation methods for the MRIO-tables in modern software, R. The quality assurance
process of different parts of the project is ongoing with the ambition to involve also
international expertise. There are good prospects that the next phase of the project will
imply a substantial step forward in the implementation of a new generation of statistical
products within Statistics Sweden providing deepened support for future policy analyses.