Eviota for Music: Methodological Case Study and Implementation Notes
An Open Policy Analysis–compliant working document prepared within the Open Music Europe (OpenMusE) project
Antal, Daniel 
Disclaimer
This document was created in the project that has received funding from the European Union’s Horizon Europe, research and innovation programme, under Grant Agreement No. 101095295.
This document has been prepared by OpenMusE project partners as an account of work carried out within the framework of the EC-GA contract no 101095295. Any dissemination of results must indicate that it reflects only the author’s view and that the Commission Agency is not responsible for any use that may be made of the information it contains.
Neither Project Coordinator, nor any signatory party of OpenMusE Project Consortium Agreement, nor any person acting on behalf of any of them:
- makes any warranty or representation whatsoever, express or implied, (i). with respect to the use of any information, apparatus, method, process, or similar item disclosed in this document, including merchantability and fitness for a particular purpose, or (ii). that such use does not infringe on or interfere with privately owned rights, including any party’s intellectual property, or (iii). that this document is suitable to any particular user’s circumstance; or
- assumes responsibility for any damages or other liability whatsoever (including any consequential damages, even if Project Coordinator or any representative of a signatory party of the OpenMusE Project Consortium Agreement, has been advised of the possibility of such damages) resulting from your selection or use of this document or any information, apparatus, method, process, or similar item disclosed in this document.
This document has not been reviewed, amended, or approved by the European Commission or the European Research Executive Agency. It therefore does not constitute an approved EC deliverable.
This document is made in accordance with Horizon Europe transparency requirements and the Open Policy Analysis (OPA) guidelines to ensure traceability, reproducibility, and a clear audit trail of the analytical work performed.
No claims of completion, endorsement, or policy validity beyond the documented scope are made.
Executive Summary
This document presents a methodological case study and implementation notes for Music Eviota, an analytical framework developed within the Horizon Europe Open Music Europe (OpenMusE) Research and Innovation Action (Grant Agreement No. 101095295).
The work documents how accounting, administrative, and statistical data can be integrated through a federated data space—the Open Music Observatory—to support economic, environmental, and sustainability analysis in the cultural and creative sectors. The emphasis is on methodological validation, standards alignment, and reproducible workflows, rather than on the delivery of a finished product or statistically representative empirical results.
Within the Horizon action, the contribution focused on validating interoperability between micro-level financial records and European statistical and sustainability frameworks (ESA 2010, Input–Output tables, and CSRD-aligned ESG concepts). Empirical demonstrations draw on limited, sensitive, or synthetic data, and on validation use cases in music and film, in order to preserve auditability and confidentiality.
This document should therefore be read as a research and infrastructure outcome: a transparent, standards-based foundation intended to support future innovation actions, policy uptake, and sectoral reuse beyond the formal end of the Horizon project.
Introduction
Music Eviota is a research-driven data space and analytical framework developed to support sustainability, economic, and policy analysis in the European cultural and creative sectors, with a primary focus on music and a closely aligned application to film and audiovisual production.
The work documented in this report was carried out under a Horizon Europe Research and Innovation Action. While product funding within the action was limited, the project enabled substantial methodological validation, cross-sector testing, and alignment with European statistical, accounting, and sustainability-reporting standards. The emphasis of the Horizon work was therefore not commercial deployment, but the development and validation of a reusable data-sharing and analytical infrastructure.
Scope and positioning
The central objective of the Horizon action was to design and validate a federated data sharing space capable of serving multiple applications interactively, rather than a single-purpose tool. This data space is implemented through the Open Music Observatory (OMO), which functions as the primary integration layer for music-sector data and as a serves as a reference implementation for film-related use cases. 12.
Within this context, Music Eviota serves as an analytical application layer built on top of the Observatory’s semantic and governance infrastructure. It enables the reuse of the same underlying data for multiple policy-relevant purposes, including sustainability reporting, economic impact analysis, and benchmarking.
The system is explicitly designed to support the full scope of CSRD-aligned ESG analysis, including:
Environmental topics (E1–E5), extending beyond greenhouse gas emissions to pollution, resource use, and ecosystem impacts;
Social topics (S1–S4), including own workforce, value-chain labour conditions, and participation;
Governance topics (G), insofar as they can be operationalised through accounting, administrative, and statistical data.
This approach differentiates Music Eviota from sector-specific carbon calculators or single-indicator tools: rather than producing isolated metrics, it embeds music and film activities into official statistical and accounting frameworks, allowing environmental, social, and economic impacts to be calculated consistently and validated against national accounts.
Central role of the Open Music Observatory
In the Horizon context, the Open Music Observatory is not an ancillary output but the core research infrastructure developed within the action. The Observatory provides:
a semantic layer that connects heterogeneous data sources (accounting records, surveys, registers, statistical tables);
shared vocabularies and identifiers that enable multilingual and cross-country interoperability;
governance mechanisms that allow data to remain with its owners while still being reused analytically.
Within the project, the following research and development objectives were achieved:
development of data models that connect IFRS-style accounting data with ESA 2010 national accounts and satellite account methodologies;
placement of benchmarked music-industry datasets into the Open Music Observatory, making them reusable across countries and applications;
extension of the open-source
iotablessoftware library to accommodate updated Eurostat SDMX vocabularies and classifications;advancement of the non-public
eviotaanalytical library to link micro-level financial accounts with Input–Output tables, including improved linguistic coverage in English, Italian, and Hungarian;scaling of the overall approach with a clear VSME focus, ensuring that the methods remain usable by small and medium-sized cultural organisations with limited technical capacity.
Relationship between music and film use cases
Although the Open Music Observatory is music-centred, film and audiovisual production play a crucial methodological role. Film production provides a highly structured, project-based accounting environment that is particularly well suited for testing satellite decompositions, tax-related analyses, and environmental extensions of Input–Output tables.
Our work greatly benefitted from the parallel work in audiovisual case studies, because audiovisual productions often have far mar more advanced enterprise resource planning tools, and can provide far more granual test data, which then can be simplified for lower granularity, more microenterprise like music data.
The Hungarian Motion Picture Data Space, which was entirely built from the modules developed for the OMO, therefore functions as a validation and stress-testing environment for methods that are subsequently applied to music-sector data through OMO. This dual focus ensures that the resulting framework is robust, policy-relevant, and transferable across cultural domains.
In summary, this report documents not a single software product, but the outcome of a Horizon Europe research effort to build and validate a federated, standards-based data space for music and film—one that enables ESG analysis, economic impact assessment, and policy evaluation to be performed in a way that is transparent, reproducible, and aligned with European statistical and sustainability-reporting frameworks.
Horizon Context and Objectives
This report documents research conducted under a Horizon Europe Research and Innovation Action, situated at the intersection of data governance, sustainability reporting, and cultural and creative sector policy. The action addressed a structural gap identified in European cultural data infrastructures: while extensive statistical, administrative, and accounting data exist, they remain fragmented across institutions, formats, and legal domains, limiting their reuse for policy-relevant analysis.
Within the Horizon framework, the primary objective of Reprex was not the delivery of a single-purpose software product, but the development and validation of a reusable data sharing space capable of supporting multiple analytical applications. This objective aligns with Horizon Europe priorities on interoperability, reuse of existing data, and the development of trusted data infrastructures rather than isolated tools.
Overall objective
The overarching objective of the action was to design and test a federated data sharing space that enables micro-level cultural sector data to be consistently connected with European statistical systems and sustainability reporting frameworks. The focus was on enabling interaction between:
company-level accounting data prepared under IFRS-like principles,
national accounts compiled under ESA 2010,
sectoral and thematic satellite accounts,
and sustainability reporting requirements under the CSRD framework.
Rather than producing new primary datasets at scale, the project focused on making existing data interoperable and analytically reusable across domains.
Specific objectives
Within this overall goal, the Horizon action pursued the following specific objectives:
To develop formal data models that link financial accounting data with ESA-based national accounts and satellite accounting methodologies, enabling consistent reclassification and aggregation across statistical layers.
To establish the Open Music Observatory as a central data integration and governance layer for the European music sector, serving both as a research infrastructure and as a reference implementation for other cultural domains.
To place benchmarked music-industry datasets into the Observatory, allowing cross-country comparison and reuse for economic, social, and environmental analysis.
To extend the open-source iotables software library to accommodate updated Eurostat SDMX vocabularies and classifications, ensuring compatibility with current and future releases of official Input–Output tables.
To advance the non-public eviota analytical library so that micro-level financial records can be systematically connected to Input–Output frameworks, including improved multilingual processing in English, Italian, and Hungarian.
To test the scalability and applicability of the approach with a focus on very small and medium-sized enterprises (VSMEs), ensuring that methodological rigor does not depend on enterprise-scale data systems or proprietary tooling.
Music and film as complementary validation domains
Within the Horizon action, music served as the primary domain for infrastructure development, while film and audiovisual production functioned as a complementary validation environment. Film production budgets and ledgers are typically more structured, project-based, and time-bounded than music-sector accounts, making them particularly suitable for testing satellite decompositions, tax-related analyses, and environmental extensions.
The Hungarian Motion Picture Data Space was therefore used to validate methods developed for the music sector under conditions of regulatory oversight, including cooperation with tax authorities, tax incentive controllers, and statistical offices. This dual-domain approach ensured that methods developed within the Open Music Observatory are robust, transferable, and compatible with public-sector validation requirements.
Positioning within Horizon Europe
The action deliberately prioritised methodological soundness, standards alignment, and reproducibility over rapid productisation. While limited product funding constrained large-scale deployment, the Horizon framework enabled close alignment with European statistical guidance, sustainability reporting standards, and data governance principles.
As a result, the outcomes of the action should be understood as a validated research and infrastructure foundation. These results are intended to support future innovation actions, policy uptake by public authorities, and downstream applications by industry actors, rather than as a finished commercial product delivered within the Horizon project itself.
System architecture and innovation
Scope of the Horizon contribution
Within the Horizon Europe Research and Innovation Action, Reprex did not aim to create a new platform or redevelop core components from scratch. The contribution corresponded to approximately three person-months and focused on targeted methodological improvements, integrations, and validation activities within an existing research and software ecosystem.
The work documented here builds on:
- the Open Music Observatory as an existing data space and research infrastructure
- the mature open-source
iotablessoftware library - a prior proof-of-concept developed under a MusicAIRE microgrant
- earlier research on cultural data spaces, satellite accounts, and sustainability reporting
The Horizon contribution should therefore be understood as incremental but substantive, strengthening analytical consistency, interoperability, and multilingual applicability rather than introducing entirely new core systems.
Backend architecture and data modelling
The backend architecture consists of a data coordination and modelling layer that enables consistent linkage between company-level accounting data and macro-level statistical and sustainability frameworks.
Within the scope of the Horizon action, the following backend-related results were achieved:
- developed and refined data models connecting financial accounting variables to ESA 2010 national accounts and satellite account concepts
- formalised relationships between ledger items, NACE economic activities, and CPA products and services to support consistent reclassification
- placed benchmark music-industry datasets into the Open Music Observatory so they can be reused for cross-country and cross-sector comparison
These developments do not replace enterprise accounting systems or official statistical production systems. They provide a semantic and analytical layer that allows data produced in different organisational contexts to be reused in a consistent and auditable way.
Analytical layer and reuse of mature components
The analytical layer combines mature open-source components with matching and reconciliation workflows developed or improved in this action.
The iotables library provides the core implementation of Supply-Use and Input-Output analysis. It predates the Horizon action and is a mature research software library with multiple public releases, an established user base, and extensive automated testing. During the Horizon project, the following improvements were implemented:
- released an improved version of
iotables, aligning its outputs more closely with UN and EU methodological guidance - aligned the implementation with updated Eurostat SDMX and Fusion Registry vocabularies and classifications
- expanded automated test coverage to approximately 900 unit tests benchmarked against reference tables and equations
In parallel, the non-public eviota analytical library was developed and improved as a bridging component that connects micro-level financial records with macro-level benchmark structures. Within the Horizon action, the key developments were:
- developed matching and reconciliation workflows to connect journal entries and trial balance items to benchmark categories used in Input-Output analysis
- improved multilingual handling and classification efficiency using English, Italian, and Hungarian vocabularies
- demonstrated high coverage in ledger classification in testing, including a film production test where automated matching covered approximately 98 percent of items
- benchmarked the emissions estimation pipeline against an industry-standard tool (Albert) on comparable cases, with results matching within approximately 5 percent under equivalent spend-based intensity assumptions
These results should be interpreted as evidence of analytical feasibility and operational efficiency rather than as a claim of complete automation or full ESRS coverage.
Frontend architecture and interactive use
In the Horizon context, Music Eviota is the analytical application layer built on top of the Open Music Observatory. The application logic is designed to reuse the same underlying accounting and benchmark data for multiple outputs, including sustainability indicators, economic impact views, and benchmarking.
Within the limited Horizon effort, the emphasis was on enabling interactive and repeatable analysis rather than building a full-featured end-user product suite. The focus was on:
- reducing manual effort in reclassification and reporting workflows
- ensuring traceability from ledger line to classification and benchmark linkage
- enabling iterative refinement through multilingual vocabularies and human review where needed
Innovation claim in neutral terms
The innovation contributed within the Horizon action is not a new statistical method or a new ESG standard. It is the integration and validation of a reusable workflow that:
- links micro-level accounting records to official macro-level benchmark frameworks (ESA 2010, NACE, CPA, Input-Output tables)
- supports multilingual ledger classification in realistic music and film accounting contexts
- improves auditability by keeping an explicit mapping path from source records to indicators
- remains usable for VSME contexts by relying on data that organisations already hold and can export
This contribution is best understood as a standards-aligned interoperability and analytical bridging result: a validated way to connect existing accounting data to European statistical and sustainability reporting frameworks with lower reporting friction and clearer verification paths.
Data spaces: film and music
This section describes the two closely related data spaces that structure the work reported here. In line with the Horizon Europe objectives, the Open Music Observatory functions as the primary research data space, while the Hungarian Motion Picture Data Space represents an exploitation and validation route that extends the same principles beyond the music industry.
The two data spaces share a common architectural logic, governance philosophy, and methodological foundation, but differ in scope, institutional embedding, and immediate policy use.
Open Music Observatory (OMO)
The Open Music Observatory (OMO) is the central data-sharing space developed and validated within the Horizon Europe action. It is designed as a federated research infrastructure for the European music sector, enabling the integration and reuse of heterogeneous data sources without centralising ownership or control.
OMO operates as a semantic and governance layer rather than a single database. Data remain with their original holders—such as music organisations, collective management organisations, statistical offices, research institutes, and private enterprises—but are made interoperable through shared identifiers, controlled vocabularies, and agreed mapping rules. This design follows European data space principles that prioritise federation, subsidiarity, and reuse over data consolidation (European Commission 2020; Data Spaces Support Centre 2025).
The governance logic of OMO reflects the realities of the music ecosystem:
- data are distributed across many small and medium-sized actors,
- legal bases for data processing vary (copyright, contracts, statistical mandates),
- and no single institution can realistically act as a central data controller.
By separating data custody from semantic coordination, OMO enables analytical reuse while respecting institutional autonomy and legal constraints. This approach is consistent with the European Interoperability Framework, which emphasises semantic mediation rather than schema unification as the basis for interoperability across heterogeneous domains (European Commission 2017).
Within the Horizon action, OMO served as the main route for:
- developing data models that connect IFRS-style accounting data with ESA 2010 national accounts and satellite account structures,
- hosting benchmarked music-industry datasets that can be reused for economic, social, and environmental analysis,
- integrating official statistical classifications (NACE, CPA) and environmental accounts into a reusable knowledge layer,
- supporting multilingual analytical workflows across countries.
The Observatory also functions as a shared research environment. It connects survey data, administrative data, financial records, and official statistics, enabling comparative analysis across countries and sub-sectors. This makes OMO suitable not only for sustainability reporting use cases, but also for cultural policy analysis, labour market studies, and innovation research in the music sector.
The design choices underlying OMO are documented in the accompanying Green Paper, which argues that centralisation is neither feasible nor desirable in the European music ecosystem, and that a federated data-sharing space provides a more robust and future-proof solution (Antal 2025).
Hungarian Motion Picture Data Space (MMAT)
The Hungarian Motion Picture Data Space (Magyar Mozgóképes Adattér, MMAT) applies the same architectural and governance principles as OMO to the audiovisual sector. It is not a separate research infrastructure developed under Horizon Europe, but an exploitation and validation route that demonstrates how the same data-space logic can be transferred beyond the music industry.
MMAT connects film production companies, granting authorities, tax rebate controllers, statistical offices, and sustainability analysts through a shared semantic layer. Like OMO, it does not centralise data. Instead, it provides mappings between:
- company accounting systems and production-specific cost codes,
- national statistical classifications (NACE, CPA),
- satellite account structures for cultural and creative industries,
- and environmental accounts used for emissions estimation.
Film production offers a particularly suitable environment for validating these methods. Productions are project-based, time-bounded, and subject to detailed financial oversight, especially where public subsidies or tax incentives are involved. As a result, film accounting data are often more granular and structured than typical music-industry records.
Within MMAT, these characteristics were used to:
- stress-test the matching of ledger entries to economic activities,
- validate reconciliation workflows against audited financial data,
- test the integration of accounting data with Input–Output tables and environmental extensions,
- and assess auditability under regulatory supervision.
Institutionally, MMAT operates in close interaction with public bodies. Its development and testing involved cooperation with tax authorities, tax incentive controllers, and statistical experts, ensuring that analytical transformations remain compatible with official accounting and reporting practices. This institutional embedding distinguishes MMAT from purely research-oriented data spaces and demonstrates the policy applicability of the approach.
While MMAT is specific to Hungary in its current form, its architecture is not country-specific. It reuses the same classifications, standards, and interoperability principles as OMO, making it transferable to other national contexts and other cultural sectors.
Methodology – from data to knowledge
This section describes the analytical methodology used to transform heterogeneous accounting, administrative, and statistical data into consistent economic, environmental, and sustainability indicators. The same methodology is shared by both film and music cases; differences arise only from data availability, institutional embedding, and sector-specific practices.
The methodological objective is not to generate new primary statistics, but to enable the reclassification, reconciliation, and analytical reuse of existing data in a way that remains compatible with European statistical standards and sustainability reporting frameworks.
Analytical scope and competency questions
The methodology is designed to answer a stable set of competency questions that recur across cultural policy analysis, sustainability reporting, and economic evaluation:
- fiscal and policy analysis, including tax incentives, subsidies, and public support schemes;
- economic and employment impacts, consistent with Supply–Use and Input–Output (SUT/IOT) frameworks;
- environmental impacts, linking expenditure patterns to sectoral emission and pollution intensities;
- operational and managerial analysis, including activity-based costing and benchmarking.
Answering these questions requires that the same accounting facts can be interpreted in multiple analytical frames without contradiction. This is achieved by aligning company-level records with official statistical classifications and satellite account logic.34
The Handbook on Supply and Use Tables and Input-Output Tables with Extensions and Applications defines the international standard for compiling IOTs and SUTs (Mahajan et al. 2018). It also establishes the procedure for adding environmental and satellite extensions, which underpins our work with pollution intensities and GVA multipliers. Our software libraries followed these implementation guidelines and contain unit tests to see if they work as requested by the statistical authorities.
Multipliers are treated in depth in Input-Output Multipliers – Specification sheet and supporting material (d’Hernoncourt, Cordier, and Hadley 2011), which provides specifications for tracing direct, indirect, and induced effects. This is especially relevant when comparing industry-wide vs. company-level impacts.
Eurostat’s Towards a Cultural Satellite Account in Europe (Eurostat_2011_Cultural?) and UNESCO’s Culture Satellite Account report (Hara 2015) both demonstrate the feasibility of disaggregating creative industries. These works were used as good examples for our our disaggregation of J59–60 and the benchmarking of music and audiovisual services as a distinct satellite block.
Tourism offers for as a mature example. The Tourism Satellite Accounts in Europe, 2023 edition (Eurostat 2023b) and its Methodological note (Eurostat 2023a) show how sector-specific accounts can be incrementally built, reconciled with national accounts, and used for policy evaluation. Tourism is also very relevant for our work because live music festivals are key drivers of cultural tourism.
From accounting records to analytical categories
At the micro level, the methodology starts from accounting artefacts that organisations already produce:
- charts of accounts and general ledger exports;
- trial balances used for statutory reporting;
- journal entries with free-text transaction descriptions.
These artefacts are not analytically sufficient on their own. Charts of accounts reflect financial logic, not economic activity. Journal entries contain richer signals, but they are heterogeneous, multilingual, and inconsistent across organisations and software systems.
The methodological solution is to add semantic structure rather than replace accounting systems. Ledger lines are enriched with labels, codes, and flags that make them analytically interpretable, while preserving a direct, auditable link to the original records.
Our system relies on integrating company-level ledgers into the IOT framework. This is supported by the UNECE Using Administrative and Secondary Sources for Official Statistics handbook (United Nations Economic Commission for Europe 2011), which explicitly legitimises the reuse of non-statistical datasets in official statistics.
Eurostat’s High-Level Expert Group reinforced this approach in Empowering society by reusing privately held data for official statistics (Eurostat High-Level Expert Group 2022), which stresses partnerships and quality frameworks for integrating private business data.
Where AI fits (and what we mean by Inference AI)
AI is used in a narrow and auditable role to assist inference at classification time (“Inference AI”). Rule-based and statistical methods annotate free-text ledger descriptions, detect likely categories, and assign provisional labels.
All automated decisions are:
- provisional rather than final;
- scored for confidence and materiality;
- fully traceable to source data and applied rules.
Items with high financial or environmental materiality are routed to human review. The goal is not to replace accounting judgement, but to focus human attention where it matters most and to ensure that the pipeline remains reproducible, transparent, and explainable.5
Alignment with official statistical frameworks
To ensure analytical validity, all classifications are aligned with official European statistical systems. The methodology relies in particular on:
- NACE classifications for economic activities;
- CPA classifications for products and services;
- Supply–Use and Input–Output tables compiled under ESA 2010;
- environmental-economic accounts, including air emissions accounts.
Expressing company-level data in these shared classification systems allows micro-level records to be embedded into macro-level analytical structures without breaking accounting identities or statistical consistency. It also enables comparison with national averages, sectoral benchmarks, and time series derived from official statistics.
All classifications are expressed using official Eurostat and SDMX vocabularies (NACE, CPA, SDMX codelists), ensuring compatibility with national accounts and environmental extensions (NACE Rev. 2: Statistical Classification of Economic Activities in the European Community 2008; “Commission Regulation (EU) No 1209/2014 Amending Regulation (EC) No 451/2008 Establishing a New Statistical Classification of Products by Activity (CPA)” 2014; “CPA 2.1: Statistical Classification of Products by Activity” 2019).
Industry × Industry tables
The main official data sources we rely on are Eurostat’s Symmetric Input–Output Tables (SIOTs).
These are available in two standard forms:
naio_10_cp1750: Symmetric IOT at basic prices, industry × industry.naio_10_cp1700: Symmetric IOT at basic prices, product × product.
Both tables are labelled with NACE Rev. 2 industries and CPA 2.1 products, and are dimensioned according to the official Euro SDMX Registry codelists (provided in all EU languages).
Eurostat / EEA vocabularies
To ensure interoperability, the system imports the following official vocabularies:
- Classifications
- NACE Rev. 2 — Statistical classification of economic activities (NACE Rev. 2: Statistical Classification of Economic Activities in the European Community 2008)
- CPA 2.1 — Classification of products by activity (“Commission Regulation (EU) No 1209/2014 Amending Regulation (EC) No 451/2008 Establishing a New Statistical Classification of Products by Activity (CPA)” 2014; “CPA 2.1: Statistical Classification of Products by Activity” 2019)
- Euro SDMX Registry codelists
SCL_IND_AVA— Industry availability side (“Codelist: Industry Availability (SCL_IND_AVA)” 2019)SCL_IND_USE— Industry use side (“Codelist: Industry Use (SCL_IND_USE)” 2019)SCL_PRD_AVA— Product availability side (“Codelist: Product Availability (SCL_PRD_AVA)” 2019)SCL_PRD_USE— Product use side (Eurostat 2019)CL_STK_FLOW— Stock/flow dimension (“Codelist: Stock or Flow (CL_STK_FLOW)” 2019)CL_GEO— Geography (countries, regions, EU aggregates) (“Codelist: Geography (CL_GEO)” 2019)CL_FREQ— Frequency (annual, quarterly, monthly) (“Codelist: Frequency (CL_FREQ)” 2019)
- Environmental extensions (EEA / Eurostat)
- Air emissions accounts — sectoral greenhouse gas intensities (CO₂, CH₄, N₂O, etc.) (Eurostat 2024)
Decomposition and satellite-account logic
Where policy-relevant activities are embedded in broader statistical aggregates, the methodology applies satellite-account logic. Aggregate industries are analytically decomposed into:
- a measured segment, based on reconciled company-level data;
- a residual segment, representing the remainder of the industry.
This approach follows established practices in cultural and tourism satellite accounts and preserves full consistency with national accounts while allowing targeted analysis of specific sectors or activities.
The disaggregation step in
eviotafollows the logic of satellite accounting. Eurostat’s Towards a Cultural Satellite Account in Europe (Eurostat 2011) and UNESCO’s culture satellite work (Hara 2015) both recommend separating cultural and audiovisual segments from broader aggregates to avoid misleading multipliers. Similarly, Tourism Satellite Accounts (Eurostat 2023b, 2023a) show how a targeted policy-relevant sector can be carved out while preserving consistency with the central accounts. By implementing this in code,eviotademonstrates how private company data and official IOTs can be combined in a reproducible, transparent, and internationally recognised way.
Outputs and analytical reuse
The outcome of the methodology is not a single indicator or report, but a reusable analytical representation of accounting data. The same enriched dataset can support:
- sustainability indicators aligned with CSRD and VSME principles;
- economic impact and benchmarking analyses;
- policy evaluation under cultural, fiscal, or environmental frameworks.
Because the methodology relies on data that organisations already collect and on official statistical standards, it remains proportionate for VSME contexts while retaining analytical rigour.
Case studies
This section presents two applied case studies that demonstrate the use of the data-space architecture and analytical methodology in practice. In line with the music-facing mandate of the Horizon action, the music-industry case is presented as the primary application. The film case is included as a secondary validation environment that supports methodological robustness and regulatory relevance.
Music industry case: MiH
MiH serves as the primary music-industry case study for the Horizon action. It represents a realistic medium-sized music enterprise operating across production, services, and touring-related activities, and therefore provides a suitable testbed for linking accounting data with economic, social, and environmental indicators.
MiH provided a complete and internally consistent set of accounting artefacts, including: - a raw ledger export (mih.xlsx),
- a chart of accounts definition (Bilancio 2024.csv),
- unique ledger descriptions (mih_ledger_descriptions.xlsx) and their coded version (mih_ledger_descriptions_coded.xlsx),
- and the official financial statements (MIH Spa - Bilancio 2024.xlsx).
These inputs are sensitive and are not present in this report or in this OPA folder.
Relevance to the music sector
The MiH case reflects several structural characteristics of the contemporary music industry:
- a high share of service-based expenditure (e.g. production services, logistics, promotion);
- extensive use of contracted and freelance labour;
- significant mobility-related activities linked to touring, events, and cross-border collaboration;
- environmental impacts that extend beyond direct energy use to transport, accommodation, and purchased services.
These characteristics make music enterprises particularly sensitive to Scope 3-type impacts and to data requests from larger partners under sustainability reporting regimes.
Data preparation and semantic enrichment
MiH’s ledger data were harmonised and enriched using multilingual declarative dictionaries and inference-assisted classification. Free-text descriptions were normalised and mapped to analytical categories, with human review applied to high-materiality items.
The classification process linked each transaction to:
- an accounting category,
- an inferred economic activity (NACE),
- a corresponding product or service category (CPA),
- and, where relevant, an activity pool suitable for activity-based costing.
The resulting dataset was reconciled with the statutory financial statements to ensure full internal consistency.
Economic and environmental linkage
Once classified, MiH’s expenditures were embedded into the Input–Output framework using national benchmark tables. This enabled:
- estimation of value-added and employment effects associated with MiH’s activities;
- calculation of environmental impacts using sectoral intensity factors derived from environmental-economic accounts;
- comparison of MiH’s intensity profile with national and sectoral averages.
Environmental analysis was not limited to greenhouse gas emissions. Where data allowed, additional pollutants and resource-related indicators were included, reflecting the broader E1–E5 scope of sustainability reporting frameworks.
Film industry validation
The film case is included as a secondary validation environment rather than a primary application domain. It draws on work conducted within the Hungarian Motion Picture Data Space and focuses on service film productions subject to public incentives and regulatory oversight.
Film production data offer several methodological advantages:
- project-based accounting with clear temporal boundaries,
- detailed cost coding linked to production activities,
- external validation through audits and tax incentive controls.
Within this context, the analytical workflow was tested against film production ledgers used in tax shelter and subsidy administration. The objectives were to:
- validate the reconciliation of ledger data with official statistical classifications;
- test decomposition of aggregate audiovisual industries into measured and residual components;
- assess auditability and traceability under regulatory scrutiny.
Results from the film validation confirmed that the same methodological approach used in the music case can operate under stricter control environments, reinforcing confidence in its robustness.
Complementarity of the cases
Together, the two cases demonstrate the transferability of the approach across cultural domains. The music case provides scale, diversity, and relevance to VSME contexts, while the film case provides depth, structure, and regulatory validation.
This complementarity supports the central claim of the Horizon action: that a federated data-space approach, grounded in official classifications and proportionate analytical methods, can serve multiple cultural sectors without sector-specific redesign.
Validation and institutional testing
A central requirement of the Horizon Europe action was that the proposed data-space and analytical workflow be credible not only in research terms, but also in institutional and regulatory environments. Validation therefore focused on testing the methodology under conditions of public-sector oversight rather than on user-facing pilots.
Institutional settings
The workflow was tested in cooperation with institutional actors involved in the governance of film and cultural production data, including:
- tax shelter controllers responsible for verifying eligible costs and compliance of audiovisual productions;
- the national tax authority, in the context of reconciling company-level accounting data with fiscal reporting and audit practices;
- statistical office experts, focusing on consistency with national accounts, Input–Output tables, and satellite account methodologies.
These environments impose strict requirements on traceability, documentation, and methodological clarity, and therefore provide a robust validation context.
Purpose of validation
The institutional testing served three closely related purposes.
First, reconciliation. Company-level ledgers processed through the system were reconciled with statutory financial statements, tax declarations, and aggregate statistical totals. This ensured that analytical reclassifications preserved accounting identities and did not introduce inconsistencies relative to officially reported figures.
Second, auditability. All transformations from source data to analytical outputs were required to be inspectable and reproducible. Validation focused on whether an external reviewer could trace each analytical result back to original ledger entries, understand the applied classification rules, and verify the use of official benchmarks and extensions.
Third, avoidance of black-box ESG. Institutional partners explicitly raised concerns about opaque ESG tools that generate indicators without clear links to underlying data. The validation therefore tested whether sustainability-related outputs could be explained in terms of accounting records, official statistical classifications, and documented assumptions, rather than relying on proprietary or unverifiable models.
Validation outcome
The testing confirmed that the workflow meets core public-sector expectations:
- analytical results remain aligned with official statistical and accounting frameworks;
- automated steps are constrained, documented, and subject to human oversight;
- the same accounting data can be reused for fiscal, statistical, and sustainability purposes without duplication.
This institutional validation is a key credibility asset of the project and distinguishes the approach from single-indicator calculators or purely commercial ESG tools.
The reuse of company ledgers follows established guidance on integrating administrative and privately held data into official statistics (United Nations Economic Commission for Europe 2011; Eurostat High-Level Expert Group 2022).
Contribution to Horizon Europe objectives
The outcomes of the action contribute to several core Horizon Europe objectives through alignment rather than through the creation of new regulatory or methodological frameworks.
European Green Deal
The project supports Green Deal objectives by enabling the integration of environmental indicators into economic and sectoral analysis using official environmental-economic accounts (Eurostat 2024). By linking company-level spending data to Eurostat air emissions accounts and Input–Output tables, the approach supports consistent estimation of upstream environmental impacts and enables evidence-based assessment of transition pathways.
European data spaces
The Open Music Observatory and its extension to the Hungarian Motion Picture Data Space implement the principles articulated in the European Data Strategy: federation over centralisation, reuse of existing data, and interoperability through shared semantics. The action demonstrates how cultural-sector data can participate in common European data spaces while respecting legal and organisational autonomy.
Cultural and creative sectors
The action addresses a well-documented gap in cultural data infrastructures by providing a standards-aligned way to connect creative-sector accounting data with national statistics and sustainability frameworks. This supports cultural policy analysis, funding evaluation, and cross-country comparison in sectors that are otherwise difficult to analyse using conventional industrial statistics.
SME and VSME inclusion
A core design principle was proportionality. The workflow relies on data that organisations already produce for accounting and compliance purposes and does not require dedicated sustainability data collection systems. This makes the approach accessible to small and very small enterprises, aligning with Horizon Europe’s emphasis on inclusive innovation and SME participation.
Evidence-based policymaking
By ensuring consistency with official statistical frameworks and by validating the workflow in institutional settings, the action supports evidence-based policymaking. The same data can be reused for fiscal analysis, economic impact assessment, and sustainability reporting, reducing fragmentation and improving the comparability of results across projects and years.
Limitations and next steps
The work reported here is subject to clear limitations.
First, funding constraints. The Horizon contribution corresponded to approximately three person-months and did not allow for large-scale deployment or extensive user-interface development. The focus was therefore on methodological validation rather than full automation or market-ready tooling.
Second, partial automation. While classification and reconciliation workflows achieved high coverage in testing, human oversight remains essential for ambiguous or material items. This is a deliberate design choice aligned with auditability requirements, but it limits the degree of end-to-end automation.
Third, scope of indicators. Environmental analysis focused primarily on air emissions and intensity-based estimates. Social and governance indicators were addressed conceptually and through accounting-linked proxies, but further work is needed to extend coverage across additional S and G dimensions.
Future steps include scaling the approach to additional countries, extending multilingual support, enriching social and governance indicators, and exploring closer integration with emerging European data-space services. These steps are envisaged as follow-on innovation or deployment actions building on the validated research foundation established in this Horizon Europe project.
References
Footnotes
European approaches to sectoral data spaces emphasise federated sharing with shared semantics rather than centralised data lakes.↩︎
European Interoperability Framework guidance on portability, semantic alignment, and governance for cross-organisation data use. .↩︎
Satellite account methodology for re-classifying domain-specific activities into national accounts without breaking core consistency.↩︎
Double materiality under EU sustainability reporting: financial materiality for investors and environmental/social materiality for stakeholders.↩︎
Human-in-the-loop classification, quality scoring, and audit trails for AI-assisted decisions in accounting and statistics.↩︎