ECA Report: Horizon 2020 - billions of euros paid for promises

The European Union funded research with the promise that ideas would turn into products, businesses would grow, and public investment would drive competitiveness. However, the European Court of Auditors (ECA) found that, in reality, funds were concentrated among a small group of beneficiaries-firms whose financial performance fell well below projections. The Court also identified instances where pre-existing patents were counted toward the program's performance indicators, as well as a monitoring system that fails to track the actual economic outcomes of the projects. According to the audit report published yesterday by the ECA, while funding was indeed granted, its translation into commercial success remained limited-and was not sufficiently verified. The audit does not uncover fraud or the misappropriation of funds, nor does it classify the entire funding as a loss; rather, it reveals significant vulnerabilities in the use of public money. European institutions financed projects for which economic prospects were not consistently or accurately assessed, and subsequently failed to systematically monitor whether the initial promises had materialized. In some instances, the gap between business plans and reality is vast. In others, the very assessment of a program's performance is compromised by incomplete data, inaccurate reporting, and the use of indicators that incorporate results achieved prior to the funding.

For example, an Italian SME included in the sample examined by the ECA auditors-it should be noted that the report published yesterday presents these cases without disclosing the names of the specific companies or entities involved-provides one of the most striking examples. By comparing the final business plan against financial statements, the Court found that the actual number of employees stood at 54% of the projected figure in 2021, 57% in 2022, 37% in 2023, and only 25% in 2024. Revenues fell even further short of the projections; the Court's auditors noted that "actual revenues were well below estimates, amounting to approximately 15% of the initial projections for 2021, 4% for 2022 and 2023, and just 1% for 2024." In Spain, another SME had total assets of euro1.5 million in 2015; its business plan-included in that year's grant agreement-projected that assets would reach euro80 million by 2019. The auditors' finding: "However, by the end of 2019, the same enterprise's financial statements reported total assets of only euro4 million.” The business had grown, but it fell short of the projection presented in the funding documentation by euro76 million.

The report does not state that failing to meet forecasts constitutes a breach of contractual obligations. Research entails risk, and a business plan is no guarantee of success. The Court further notes that, for at least three of the 13 single-beneficiary SMEs examined-which provided financial statements for at least some of the years in question-the initial plans had significantly overestimated economic performance (in terms of revenue or employment) by at least one-third.

Moreover, the auditors observed instances where overly optimistic estimates, presented in a funding application, were revised downward in subsequent applications. Their conclusion points to the very control that was lacking: "A systematic assessment by the Commission at the end of a project could strengthen the incentive for participants to develop more robust and realistic proposals and business plans.”

• Ten research organizations, participating in 1,700 of the 6,500 funded projects

These examples are part of a broader picture. Horizon 2020, the European research and innovation programme for the 2014-2020 period, had a budget of approximately euro79 billion, including Euratom funding. The "Industrial Leadership' priority, which was examined by the European Court of Auditors (ECA), received approximately euro17 billion. Of this amount, euro13.6 billion was allocated to industrial and enabling technologies, euro2.8 billion to access to risk finance, and approximately euro0.6 billion to innovation in SMEs. The audit covered the 2014-2025 period and included an examination of 33 completed projects, visits to 25 SMEs across five Member States, and an analysis of the selection, monitoring, and evaluation systems.

The distribution of funding illustrates the extent to which resources were concentrated. Within the scope analyzed-covering generic and industrial technologies as well as innovation in SMEs-approximately 14,500 legal entities participated in projects, with 76% of these being for-profit entities. However, half of the support went to about 550 entities-representing only 3.8% of the participants-consisting primarily of non-profit organizations, research bodies, and educational institutions.

"For instance, the 10 research organizations that received the largest amounts of support participated in over 1,700 out of 6,115 projects (28%). These organizations often benefited from the highest funding rate (up to 100%), which means that this support failed to act as an incentive for attracting private investment in research and innovation," the Court notes.

The report does not disclose the names of these ten organizations, nor the individual or total amounts they received. While such a concentration of funds does not in itself constitute an irregularity, it raises a legitimate question-within the context of a program intended to strengthen industrial standing-regarding the funding's capacity to mobilize private investment and generate commercial outcomes.

The project selection process also lacks sufficient safeguards. Among the 30 evaluation reports examined by the ECA, the projected economic impact and the potential for business expansion were assessed in only three instances. In 15 of the reports, the scores were based on criteria such as turnover and the number of employees, yet no explanation was provided as to how these figures were determined. For another 12 cases, the overall impact score was not supported by sufficient justification regarding individual indicators or the calculation method.

"In the absence of a clear audit trail demonstrating how evaluators assessed the expected impact and the validated proposals, the scores may appear arbitrary, and their reliability questionable," the auditors warn. While this finding concerns the specific sample examined, it undermines the credibility of the mechanism intended to channel funds toward promising projects.

• Small SMEs excluded from the outset from EIB InnovFin funding

The report also identifies instances of participation that did not directly contribute to developing innovation capacity. One SME from Hungary was involved in at least 12 projects, solely for services such as proposal drafting, market research, and financial forecasting. "These activities cannot be considered to enhance innovation capacity," the Court observes. While this situation is not classified as fraud, it raises concerns regarding the funding of participants whose contributions do not align with the objective of strengthening innovation within enterprises.

Beyond project grants, the InnovFin instruments reveal further shortcomings in targeting and control. Products managed by the European Investment Bank primarily reached large-scale projects and mid-cap enterprises. Minimum loan thresholds-often set at euro25 million-effectively excluded many small firms from the start. According to the EIB's operational report from December 2024, cited by the auditors, SMEs accounted for 47% of the final recipients but received only 14% of InnovFin funding: euro1.42 billion out of euro10.5 billion.

These figures refer to the financing of InnovFin products managed by the EIB and should not be confused with the euro2.8 billion European budgetary allocation for access to risk finance.

Regarding InnovFin products managed by the European Investment Fund (EIF), essential data on the outcomes of the support are missing. The SME Guarantee facility reached 37,440 final recipients. Approximately 42% of the financial support was linked to the criterion of developing innovative products, processes, or services. However, the Court notes that the EIF did not track whether the supported SMEs successfully commercialized or disseminated their research results, even though the delegation agreement required such monitoring.

• Patents existing prior to the submission of innovation projects

The funded research was not without results. Ten of the 13 single-beneficiary SMEs visited successfully completed a technology transfer, moving from concept to a validated innovation. Yet, seven of these 13 reached the commercialization stage without achieving market adoption of their technological outcomes. Among the obstacles encountered, seven cited a lack of funding, six pointed to low demand, and four identified a flawed market strategy.

There are also documented commercial successes. An Italian SME developed an automated machine for the inspection of lyophilized pharmaceutical products, achieved sales and exports, and grew into a large enterprise. This description corresponds to the LYO-CHECK project, for which the report's annex indicates an EU contribution of approximately 1.72 million euros. The example demonstrates that the mechanism could work. The problem is that success was not systematically pursued, and projects with commercial potential were not consistently identified for follow-up support.

Conversely, the picture of performance could be distorted by the very indicators used. For projects related to generic and industrial technologies, only 147 out of the 642 patent applications reported by 2025-approximately 23%-had been filed after the projects began. Of the 261 granted patents, only 62-about 24%-originated from applications filed after the start of the projects; the remainder consisted of pre-existing patents. "These IPRs [intellectual property rights] had no connection to the funding provided under Horizon 2020, yet they accounted for a substantial portion of the indicator values," the cited report notes. The auditors point to methodological issues that led to a "systematic overreporting." At the same time, other relevant outcomes-such as those protected by trade secrets or software licenses-were not adequately captured by the monitoring process.

Incomplete and inconsistent reporting was identified in 22 of the 33 projects examined. In five of the 18 collaborative projects, the number of "innovative enterprises" reported exceeded the total number of firms and SMEs participating in the consortia. In three of the 15 single-beneficiary projects, the same firm was counted twice. The data were self-reported, with no systematic verification by the Commission.

Furthermore, the assessment of economic benefits relied on models whose underlying assumptions could favor overly optimistic conclusions. The Court points out that macroeconomic models assumed the supported innovations would achieve "full commercial success,” despite a lack of information regarding the outcomes for participants following the receipt of funding.

"To become a true industrial leader, funding research is not enough. The EU must consistently ensure that public investments go the full distance, translating into market-ready innovation, stronger private-sector participation, and sustainable economic growth. However, this has not always been the case,” stated Petri Sarvamaa, the Court member responsible for the audit, adding: "A significant gap persists between research and commercialization-even for research that is closest to the market. The next generation of European research and competitiveness programs must bridge this divide.”

This warning comes ahead of a major new budgetary commitment: the European Competitiveness Fund and the Horizon Europe program, proposed for the 2028-2034 period, have a combined planned budget of approximately euro409 billion (in current prices). Consequently, the European Court of Auditors (ECA) calls for the consistent application of impact criteria, support for market entry and business scaling, long-term monitoring of beneficiaries, and more robust evaluation methods.

The issue highlighted by the audit is that public funders must be able to distinguish between justified technological risk, unrealistic commercial promises, and outcomes reported via indicators that bear no relation to the funds actually granted. While the European Union has funded the development of various technologies-achieving some successes along the way-the audit underscores the need for greater discernment. However, regarding a key part of the economic promise, it lacked the necessary evidence and checks to convincingly state what was actually purchased with taxpayers' money.