Introduction

The European Union is a significant source of funding for media and journalism in and beyond Europe. Through development-cooperation instruments, research programmes, culture and citizenship funding and its own communication budgets, it supports independent media, press freedom, media literacy, journalism research and related work. Understanding where that money goes is a precondition for asking how it bears on the sustainability of independent media at a moment when the economics of the sector are under acute strain.

This study maps EU media- and journalism-related contracted funding recorded in the EU Financial Transparency System (FTS) over the eleven years from 2014 to 2024. It is built from a systematic keyword search of the FTS subject field, screened record by record, and analysed at the level of the individual beneficiary entry, so that each beneficiary listed against a multi-partner award is counted separately with its own contracted amount. It describes the structure of the recorded funding (its scale, its evolution, its geography, the programmes and departments through which it flows, the organisations that receive it and its thematic composition) and distinguishes support-oriented funding from the procurement of communication and analytical services for the EU itself. Components may not sum precisely to totals because of rounding.

The report is descriptive rather than evaluative: it does not assess the quality, delivery or impact of individual projects, nor judge allocation choices.

Two boundaries define the dataset. It covers direct and indirect management: most expenditure under shared management lies outside FTS, although the dataset nevertheless retains the limited number of ERDF-, EAGF- or other shared-management-labelled records that appear directly in FTS; and it records contracted amounts, not money ultimately paid. Within those bounds it is offered as a shared, reproducible baseline against which the effects of EU support can subsequently be tested.

Executive summary

Between 2014 and 2024 the EU Financial Transparency System (FTS) records €1,061.5 million in beneficiary-level contracted amounts for media- and journalism-related activity under direct and indirect management. Analysed at the level of the individual beneficiary entry, the dataset contains 6,103 beneficiary entries naming 2,166 entities in 105 countries. Of these, 4,657 entries carry a positive beneficiary-level contracted amount, covering 1,798 entities in 84 countries; the remaining 1,446 entries identify listed consortium or framework participants for which FTS records no beneficiary-level amount. Except for the complete-population figures just given, all subsequent counts of funded records, entities and countries, and all median calculations, are based on positive-amount records; the zero-value entries are retained in the dataset and reported separately: excluding them does not affect the monetary totals. All amounts are nominal and represent amounts initially contracted, which may differ from amounts ultimately paid.

Each positive-value record is classified analytically according to the transaction represented in its full subject description. The classifier first asks whether the record identifies a funded grant or research action, using a grant-agreement or project number, an acronym-and-title structure, explicit grant or call language and consortium context as corroborating signals; no single identifier or keyword is decisive. A record is classified as procurement where the subject describes a framework or specific contract, task order, staffing assignment, audit, evaluation, event organisation, media-buying order, or the provision of communication, monitoring, commissioned research or IT services. A project identifier or acronym is treated as evidence of support only where the full subject describes a funded action; it does not by itself override explicit framework-contract, task-order or service-purchase language.

On this basis the dataset divides into support-oriented funding of €758.3 million (2,897 records, 1,601 entities) and procurement of services of €303.1 million (1,760 records, 237 entities). Each procurement record is assigned to one primary service category, making the four mutually exclusive: advertising and media-buying (€114.2 million), communication and events (€133.0 million), media monitoring, commissioned studies and opinion research (€53.8 million) and information technology (€2.1 million). Entity counts by category are not additive: 40 entities hold records in both categories.

The support-oriented view is a better approximation of EU support for media- and journalism-related activity than the full dataset, because it excludes procurement for the EU’s communication, analytical and operational needs. It is not, however, a measure of direct newsroom support alone: the €758.3 million category also includes universities, research projects, audiovisual and film activity, media technology and other media-related interventions. The median positive-amount record is €68,480 across the full dataset and €90,577 among support-oriented records. Many positively funded beneficiary allocations within multi-partner projects are modest, while a small number of very large records raise the mean well above the median.

Recorded funding rose from €37.5 million in 2014 to €186.2 million in 2024, with a step-change after 2021: the annual total jumped 66.9% between 2021 and 2022. Funding is concentrated among beneficiaries in France, Belgium and Germany (51.8% of the full dataset), though far less so in the support-oriented view (35.8%). It flows principally through external-action and research instruments, and its single largest recipient is a media-buying agency contracted to place advertising on the Commission’s behalf.

Methodology

Source

The analysis draws on the European Union’s Financial Transparency System (FTS), the Commission’s public record of the beneficiaries of EU budget implementation under direct and indirect management, which also includes legacy European Development Fund (tenth and eleventh EDF) records. FTS publishes, for each award, the year, a budget reference, the beneficiary’s name, VAT number and country, two monetary values (the beneficiary’s contracted amount and the commitment amount), the programme name, the subject of the grant or contract, the responsible department, and project start and end dates. All 12 fields are retained in the working dataset. The period covered is the 2014 to 2024 financial years, and the extraction was carried out in February 2026.

Because FTS operates under direct and indirect management, the dataset does not include funds disbursed under shared management (the European Regional Development Fund, the European Social Fund Plus, or the Common Agricultural Policy) except where those appear in FTS records directly. Media-related funding routed through national managing authorities under those instruments therefore falls outside this analysis.

Search strategy: keywords and semantic families

FTS is not organised by sector, so media-related records cannot be drawn from a pre-existing category, which is the reason why we conduct this series of studies in the first place. They were identified by keyword matching against the free text “subject of grant or contract” field, organised into five semantic families:

Master term — media. Core sector — journalism, journalist, journalists, newsroom, news media, press freedom, broadcasting, audiovisual. Content and function — investigative journalism, investigative reporting, fact-check, community media, independent journalism, public interest journalism. Information and democracy — disinformation, misinformation, fake news, information manipulation, media literacy, information literacy, media pluralism, media freedom, freedom of expression, freedom of information. Support and policy — media support, media development, safety of journalists, media viability, media capacity building.

The distinct terms above number 30. Several terms were searched in additional language or spelling variants; 27 exports were combined, although per-pass result counts were not preserved. One methodological caution applies to the master term in particular: matching “media” as a substring also captures unrelated words (mediation, mediated, medial and the Spanish mediante) and a word-boundary rule, matching “media” and its genuine compounds but not those stems, is the appropriate basis. The screening described below applies that correction.

Each search returned every FTS entry whose subject line contained the term. Results were exported separately, tagged with the originating keyword, and then merged. Some support-and-policy terms returned few or no results; they are retained in the record of the search so that the coverage of the query is fully documented.

Consolidation, deduplication and relevance scoring

The keyword exports were combined into a master file. Because a single award can match several keywords and appear under more than one budget reference, the raw merge contained substantial duplication. Deduplication is performed on the combination of budget reference and entity key, never on budget reference alone, so that each beneficiary listed against a multi-partner award is retained as its own record.

Each entry was then assigned a relevance class. Entries with an explicit focus on media, journalism, information integrity, freedom of expression, media literacy or closely related institutional reform were marked highly relevant; those with a substantial media component inside a broader project, relevant; those using “media” generically or in ambiguous wording, uncertain; and the remainder off-topic. This classification provided a conservative core of clearly media-related awards while isolating an uncertain group for closer manual examination.

Manual verification

The classified dataset was then screened to remove keyword false positives. Because the master term “media” was matched as a substring, and because the word “media” itself carries unrelated meanings in scientific and technical usage, a semantic rule rather than a simple word-boundary rule is required: a record was removed only where every occurrence of a matched term fell outside any media, journalism, press or broadcasting sense. Two families of false positive were identified. The first, lexical, arises where the string “media” sits inside an unrelated word: peace, diplomatic or legal mediation; biomedical or technical uses of “mediated”, “mediating”, “mediator” or “medial”; and Spanish “mediante”. The second, semantic, arises where the genuine word “media” is used in a physical, chemical or biological sense (porous media, particulate media, scattering media, culture and growing media) together with a handful of acronyms (a molecular-biology project named BROADCAST, and Galileo satellite “broadcast” records), company names (BIT MEDIA, CINE PLUS MEDIASERVICE), place names and the Spanish “medianas empresas” (small and medium-sized enterprises). Genuine media and broadcasting research, for example UHD-broadcasting and immersive-media projects, was retained or restored on individual review.

After screening and the removal of two biomedical false positives and two duplicate records, the dataset comprises 6,103 beneficiary-level award records. Records sharing the BBC’s VAT number were separated so that BBC Media Action is reported apart from the BBC’s research and technology participation. Because a number of borderline records may warrant a different decision on individual review, the total should be read as a screened, manually reviewed best estimate, open to correction.

Monetary basis

A note on terms of measurement is essential. FTS records two amounts per record. The beneficiary’s contracted amount is the amount initially contracted to the named beneficiary; over a project’s life the beneficiary may ultimately receive more, less or nothing, so it is a record of contracted funding, not of money paid. The commitment amount is the total value of the budget commitment; for multi-partner projects a single commitment covers all partners, so it must not be summed across a consortium. The two agree to the nearest euro in 52.1% of positive-amount records. This report uses the beneficiary-level contracted amount throughout, reported in nominal euros.

A note on the unit of analysis is also important. Each record is one beneficiary entry within an award, identified by a budget reference and a named organisation. Multi-partner awards therefore appear as several records. Of the 6,103 beneficiary entries, 4,657 carry a positive contracted amount; the remaining 1,446 identify listed consortium or framework participants for which FTS records no beneficiary-level amount. Counts of funded records, entities and countries, and all medians, use positive-amount records; the zero-value entries are retained in the dataset and noted where relevant, but they do not affect the monetary totals. A total of 57 records use a generic natural-person placeholder and are kept separate.

Consolidation of beneficiary names

Organisations appear in FTS under inconsistent name strings (variant spellings and acronyms appended after asterisks) so a single organisation may be recorded under several forms. To make the beneficiary tables reproducible, this report adopts a single consolidation basis: the VAT/name-keyed beneficiary entity. Records sharing the same VAT number are combined; records with no usable VAT (blank or “-”) are grouped by cleaned name; and the 57 records under the generic “natural person” placeholder are each kept separate. This is a pragmatic entity key rather than a strict legal entity register: a VAT number can occasionally span related entities or a VAT group, so a small number of manual separations are applied (for example the British Broadcasting Corporation’s research record is separated from BBC Media Action, which shares its VAT). The accompanying dataset carries the raw FTS name, the VAT number and a canonical display name so that every beneficiary figure can be reproduced and revised. On this basis the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) consolidates to €24.0 million, while the several Internews entities are reported separately by VAT.

Thematic classification

The thematic breakdown regroups terms from the search strategy into seven analytical themes matched against the subject field. It is reported as indicative rather than definitive. The categories are not mutually exclusive (an award may match more than one, so theme totals sum to more than the dataset) and they do not partition it: just over half of the dataset by value carry descriptions general enough that no specific theme can be assigned, or fall outside the seven selected themes, and these are reported as an “Other / general” group in the thematic table. The programme and departmental breakdowns, which rest on the Commission’s own budget nomenclature rather than on keyword inference, are the more reliable guide to the dataset’s composition.

Media-buying and communication contracts

A central distinction in this dataset is between support and procurement, applied at record level and according to the transaction represented in the full subject description. A grant-agreement or project number, an acronym-and-title structure, explicit grant or call language and consortium context are treated as corroborating evidence of a funded project, but no single identifier or keyword is decisive. Procurement classification is applied where the subject describes a framework or specific contract, task order, staffing assignment, audit, evaluation, event organisation, media-buying order, or the provision of communication, monitoring, commissioned research or IT services. Explicit purchase language overrides the financing programme, while service-related words appearing within the title or topic of an identifiable funded project do not by themselves make the record procurement. Support-oriented funding totals €758.3 million; procurement €303.1 million, each procurement record assigned to one primary category: advertising and media-buying (€114.2 million), communication and events (€133.0 million), media monitoring, commissioned studies and opinion research (€53.8 million) and information technology (€2.1 million).

Limitations and status

The method has documented strengths: systematic coverage of FTS direct and indirect management over the period, an explicit multi-term search strategy, and beneficiary-level detail supporting concentration analysis. The support-versus-procurement classification is an MJRC analytical judgement, not an FTS legal-status field, and remains open to correction at record level. Known service suppliers are used only to flag records for review, not to determine the outcome, so the same entity may hold both support-oriented and procurement-oriented records; and repeated combinations of entity key and normalised subject are checked for classification consistency. The accompanying dataset publishes, for every record, its transaction class, its service subtype where applicable, the basis for the decision and a manual-review flag, so that every figure can be examined and revised.

The big picture

Scale and scope

The dataset contains 4,657 positive-amount records totalling €1,061.5 million over eleven years, covering 1,798 entities in 84 countries. The median record is €68,480, and the mean is substantially higher: a small number of very large contracts sit alongside a long tail of modestly funded consortium participants. The largest single record is €19.5 million.

Temporal evolution, 2014–2024

Recorded contracted funding rose from €37.5 million in 2014 to €186.2 million in 2024. The rise was not steady: the annual total climbed from €49.4 million in 2017 to €119.0 million in 2021, jumped 66.9% to €198.7 million in 2022, then eased to €160.7 million in 2023 and €186.2 million in 2024. The eight years to 2021 averaged €64.5 million annually; the three years from 2022 averaged €181.9 million. The timing of the step-change is consistent with, but does not by itself establish, a response to Russia’s full-scale invasion of Ukraine in February 2022 and the expansion of EU commitments on strategic communication and independent media in the Eastern neighbourhood that followed.

Geographic distribution

The figures in this section describe beneficiary registration or contracting location, which is not necessarily where the funded activity took place. On a positive-amount basis: beneficiaries in France, Belgium and Germany together account for 51.8% of the full dataset. In the support-oriented view this concentration falls to 35.8%, because much of the Western European weighting reflects communication, monitoring and IT companies that contract with the EU from Brussels and Paris. Once procurement is set aside, support-oriented funding is markedly more dispersed, reaching organisations across the enlargement region, the Eastern neighbourhood, Africa and Asia. Belgium remains the largest single support location at 12.8%.

Funding programmes

After stripping budget-code prefixes and merging identical names, 82 distinct budget programmes appear. On a positive-amount basis the largest is Horizon 2020 (€146.1 million), followed by the Neighbourhood, Development and International Cooperation Instrument – Global Europe (€141.7 million) and the budget line for actions financed under the Commission’s own prerogatives (€102.0 million). Grouped by function, external-action and enlargement instruments account for 31.7% of the total, research and innovation for 23.4%, and culture, education and citizenship for 10.1%.

Responsible departments

Institutional responsibility is led by DG CNECT (€220.9 million), followed by DG NEAR (€210.6 million) and DG COMM (€179.8 million), with EACEA close behind. In total 39 department categories appear; the eight largest account for 85.3% of the positive-amount total. Legacy X-coded department labels are grouped together in the table.

Media-buying and communication services

A substantial share of the recorded funding does not constitute support-oriented funding to the media sector; it procures communication, analytical and operational services for the EU itself. On the record-level classification, procurement totals €303.1 million across 1,760 records. The largest procurement category is communication and event services, at €133.0 million. Advertising and media-buying follows at €114.2 million and is dominated by Havas.

Beyond advertising, procurement comprises communication and event services (€133.0 million), media monitoring, commissioned studies and opinion research (€53.8 million, including firms such as Kantar and Unicepta) and information technology (€2.1 million). These are legitimate EU expenditure and appear in FTS alongside grants, but they are procurement of services rather than support to the media sector. Because the classification is made at record level, a company that mainly sells services to the EU may also appear as a genuine grant partner in a research consortium.

The distinction is analytical, not a legal determination: FTS does not supply a clean grant-versus-contract field, so the classification rests on subject descriptions, programme context and procurement terminology, and remains open to correction at the level of the individual record.

Institutional concentration

Grouping by beneficiary entity on a positive-amount basis, the ten largest recipients in the full dataset hold 29.4% of the total. The largest is Havas (€101.0 million), a media-buying agency classified as procurement, followed by the European Endowment for Democracy (€49.0 million) and Kantar (€26.6 million, media monitoring). Support recipients and procurement suppliers are interleaved in the full ranking.

The support-oriented top ten is composed of media-development organisations, press-freedom bodies, broadcasters and universities: the European Endowment for Democracy, GIZ, International Media Support, BBC Media Action, Internews and the European University Institute among them. The procurement suppliers that appear in the full-dataset ranking are excluded here by classification.

Concentration among support recipients is moderate. The single largest holds 6.5%, the top ten 23.4%, and the top fifty about 45.6%. At the other end of the distribution, 1,072 support entities (around 66% by count) appear only once across the years studied.

Thematic composition

Award descriptions can be classified by theme using keyword matching on the FTS subject field. The method is approximate and the themes overlap. On a positive-amount basis the largest identifiable themes are information integrity, grouping disinformation, misinformation and fact-checking (€120.2 million); media support and development (€117.4 million); core journalism (€114.7 million); and media freedom and pluralism (€97.1 million). Records carrying none of the seven thematic flags account for just over half of the dataset by value (51.7%); this group includes both generally worded records and records whose specific subject falls outside the seven selected themes.

Journalist safety

Records whose subject contains the exact phrases “safety of journalists” or “protection of journalists” total €1,458,772 across 9 positive-value records. This is a narrow, reproducible exact-phrase indicator, not a measure of the full resources devoted to journalist safety, much of which sits within broader media-freedom programmes without separate tagging.

The largest individual awards

The 20 largest records account for about 16.0% of the positive-amount total, and the 12 shown in Table 10 for about 11.7%. The single largest is €19.5 million to the European Endowment for Democracy in 2022, for support to civil society and independent media. The largest records combine external-action grants for media support abroad with advertising and media-buying contracts.

Conclusion

What the study tells us

This mapping provides a consistent, FTS-based account of EU media- and journalism-related contracted funding over 2014–2024, analysed at beneficiary level and classified by the nature of the transaction. Read as a description of recorded contracted funding rather than of impact, several findings stand out.

The recorded field is substantial and widely distributed. FTS records €1,061.5 million in contracted amounts against 1,798 entities in 84 countries. The dataset identifies a broad population of organisations, including many modestly funded consortium participants, rather than a concentrated set of large intermediaries.

The composition matters as much as the total. Of the €1,061.5 million, €758.3 million is support-oriented (grants and support to media organisations, press-freedom bodies, broadcasters, universities and research) while €303.1 million procures services for the EU: advertising, communication, monitoring, commissioned studies and IT. The single largest recipient is not a newsroom but a media-buying agency. The support-oriented view is a better approximation of EU support for media- and journalism-related activity, though it is broader than direct newsroom support alone.

Support-oriented funding is far less geographically concentrated than headline figures suggest. Where the full dataset shows a heavy France–Belgium–Germany weighting (51.8%), the support view falls to 35.8%, because much of the concentration is procurement suppliers contracting from Brussels and Paris.

How the study can be used

The value of this mapping is as pre-impact infrastructure: a structured, documented and reproducible baseline that later work can build on. Several uses follow directly.

The support-oriented view excludes every record classified as procurement of services for the EU, on the record-level basis described in the Methodology. Because that classification is made per record, an organisation may contribute support records and procurement records to different totals.

For policy debate, the study offers a shared factual basis where consolidated figures have been lacking. It allows questions about EU media support to be posed in concrete terms: how much recorded funding reaches journalism as practised rather than the study, communication or infrastructure around it; how the balance between external-action, research and communication instruments reflects strategic priorities; whether the geography of intermediation serves the goal of media pluralism across and beyond the Union; and whether stated commitments, such as journalist safety, are matched by recorded resources. The mapping does not answer these questions, and is not intended to; it is offered so that they can be debated against evidence rather than impression, by researchers, policy-makers, experts and others, and revised as the underlying data and classification are refined.

Media Funding Series · Research report

Tracing the Flows: A Structural Analysis of EU Media and Journalism Funding in 2014–2024

By Marius Dragomir and Flora Alfiero · Media and Journalism Research Center · August 2026

DOI10.5281/zenodo.21786132 Download PDF APA BibTeX RIS Copy citation

Cite this study

Dragomir, M., & Alfiero, F. (2026). Tracing the Flows: A structural analysis of EU media and journalism funding in 2014–2024. Media and Journalism Research Center. https://doi.org/10.5281/zenodo.21786132

@techreport{dragomir2026tracing,
  author      = {Dragomir, Marius and Alfiero, Flora},
  title       = {Tracing the Flows: A Structural Analysis of EU Media and Journalism Funding in 2014--2024},
  institution = {Media and Journalism Research Center},
  year        = {2026},
  type        = {Research report},
  doi         = {10.5281/zenodo.21786132},
  url         = {https://doi.org/10.5281/zenodo.21786132}
}

TY  - RPRT
AU  - Dragomir, Marius
AU  - Alfiero, Flora
TI  - Tracing the Flows: A Structural Analysis of EU Media and Journalism Funding in 2014-2024
PY  - 2026
PB  - Media and Journalism Research Center
DO  - 10.5281/zenodo.21786132
UR  - https://doi.org/10.5281/zenodo.21786132
ER  - 

About the authors

Marius Dragomir

Author · Director, MJRC

Dr Marius Dragomir is the founder and director of the Media and Journalism Research Center, and has been a researcher and professor at the University of Santiago de Compostela and at Central European University in Vienna. He previously directed CEU’s Center for Media, Data and Society, and leads long-running research projects including the Media Influence Matrix, the State Media Monitor and the Media Capture Monitoring Report. In 2025 he won the Media and Democracy Karol Jakubowicz Award for the Media Capture Monitoring Report, recognised by the selection committee for its systematic and pioneering perspective on media ownership, regulatory frameworks and the relationship between media and politics. His studies and articles on journalism and media policy have been published and translated in more than 60 countries.

Flora Alfiero

Author

A freelance journalist based in the Netherlands, reporting on labour rights, environmental justice and university issues for outlets including The European Correspondent and Il Fatto Quotidiano.