Shares of EU funds that go towards cycling investments must be taken with a grain of salt, and the reason for this is quite simple. The EU operates through “Categories of intervention” (CIs) to see how much money has been allocated to what end. Among these, two overtly refer to cycling: 090 and 083.
Our research shows that such cycling-related CIs are frequently being used for projects that have very little, if anything at all, to do with cycling. And, reversely, that cycling projects can be found under other CIs where they should not have been included in the first place.
This report introduces a novel methodology, focusing on individual project descriptions rather than CIs, to establish a €6.8 billion preliminary estimate of EU contributions to cycling investment across 2014–2027.
The challenge: Cycling investment is hard to identify
EU regulations mandate member states to report EU-funded projects under specific Intervention Fields (IFs).
In Kohesio, from which the dataset for this study is retrieved, IFs are named “Categories of Intervention” (CIs). These are numerical codes to which a “nomenclature”, meaning a content description, is assigned. Ideally, member states should report projects under the CI/IF that better reflects their content. Nonetheless, our analysis reveals this is often not the case:
These inconsistencies make it impossible to rely on CIs alone for accurate estimates.
The solution: A hybrid approach
To address these gaps, we challenged the concept of CI, by understanding them as possible containers of cycling investment rather than solid descriptive units. We analysed 7,713 projects reported under CIs 090, 043, and 083. Using Python scripts, we applied a content analysis to project titles and descriptions, assigning Cycling Scores (CS) based on keyword detection:
This method filters out unrelated projects and provides a more reliable figure, leading to an estimate of €6.8 billion in EU contributions to cycle-related projects. The figure is yielded by the sum of CS 2, 3 and 4; meaning it factors in all projects with at least one cycling component (i.e. cycle-related projects). A narrower estimate will be shortly published in a separate policy brief.
Key findings: The inconsistency of CIs
Of the €20.7 billion in EU contributions allocated across the three CIs, €6.8 billion targets cycling related projects.
Consequently, CIs lose their meaning: rather than being a transparency measure, their inconsistency makes budget expenditure tracking increasingly complex. This inconsistency is triggered by what we define as “misreporting”. On top of reporting projects under unrelated CIs, drivers of misreporting include:
To illustrate CI inconsistency, the results for CI 090 are presented below. Individual results and discussions on misreporting across each CI are available in the report.
Two examples from the case studies: The Ericeira Intermodal Park and the Cycle Path Network in Čepin
The report illustrates Cycling Scores through 16 case studies. An example of CS0 is Portugal’s €1.1 million logged under CI 090 - cycle tracks and footpaths, which funded:
This case exemplifies how misreporting inflates cycling budgets, by including projects and ancillaries that do not relate to the nomenclature of the CI under which they are reported.
Contrarily, the development of cycle paths in the Croatian town of Čepin, categorised under CI 043, proves the existence of significant cycling investment under this CI. The project had a total budget of €1,220,314.24 and an EU contribution of €837,852.03, contributing to the development of 6.3km of cycle paths.
What needs to change
Download the report now.
Our research shows that such cycling-related CIs are frequently being used for projects that have very little, if anything at all, to do with cycling. And, reversely, that cycling projects can be found under other CIs where they should not have been included in the first place.
This report introduces a novel methodology, focusing on individual project descriptions rather than CIs, to establish a €6.8 billion preliminary estimate of EU contributions to cycling investment across 2014–2027.
The challenge: Cycling investment is hard to identify
EU regulations mandate member states to report EU-funded projects under specific Intervention Fields (IFs).
In Kohesio, from which the dataset for this study is retrieved, IFs are named “Categories of Intervention” (CIs). These are numerical codes to which a “nomenclature”, meaning a content description, is assigned. Ideally, member states should report projects under the CI/IF that better reflects their content. Nonetheless, our analysis reveals this is often not the case:
- 2014-2020: CI 090 (cycle tracks and footpaths) includes unrelated projects.
- 2014-2020: CI 043 (clean urban transport infrastructure including equipment and rolling stock) contains €3.4 billion in cycling investments, despite CI 090 being the designated category for cycling.
- 2021-2027: CI 083 (cycling infrastructure) still features projects targeted to pedestrians or with significant nonrelated elements.
These inconsistencies make it impossible to rely on CIs alone for accurate estimates.
The solution: A hybrid approach
To address these gaps, we challenged the concept of CI, by understanding them as possible containers of cycling investment rather than solid descriptive units. We analysed 7,713 projects reported under CIs 090, 043, and 083. Using Python scripts, we applied a content analysis to project titles and descriptions, assigning Cycling Scores (CS) based on keyword detection:
- CS 0: no cycling investment
- CS 1: only pedestrian investment
- CS 2: cycling and non-cycling hybrid investment
- CS 3: cycling and pedestrian hybrid investment
- CS 4: only cycling investment
This method filters out unrelated projects and provides a more reliable figure, leading to an estimate of €6.8 billion in EU contributions to cycle-related projects. The figure is yielded by the sum of CS 2, 3 and 4; meaning it factors in all projects with at least one cycling component (i.e. cycle-related projects). A narrower estimate will be shortly published in a separate policy brief.
Key findings: The inconsistency of CIs
Of the €20.7 billion in EU contributions allocated across the three CIs, €6.8 billion targets cycling related projects.
Consequently, CIs lose their meaning: rather than being a transparency measure, their inconsistency makes budget expenditure tracking increasingly complex. This inconsistency is triggered by what we define as “misreporting”. On top of reporting projects under unrelated CIs, drivers of misreporting include:
- Multi-CI reporting: Eight member states report the same project under multiple CIs. This inflates cycling investment figures by factoring in nonrelated yet significant ancillaries.
- Poor data quality: Descriptions are often mistranslated, incomplete, or use improper terms (e.g., “footbridge” for cycle bridges).
- Regulatory gaps: The absence of a pedestrian-specific CI in the 2021–2027 financing period forces misreporting into CI 083.
To illustrate CI inconsistency, the results for CI 090 are presented below. Individual results and discussions on misreporting across each CI are available in the report.
Two examples from the case studies: The Ericeira Intermodal Park and the Cycle Path Network in Čepin
The report illustrates Cycling Scores through 16 case studies. An example of CS0 is Portugal’s €1.1 million logged under CI 090 - cycle tracks and footpaths, which funded:
- A large parking infrastructure for cars and coaches (209 spaces)
- A bus terminus
- Traffic reorganization in the area
- Only five bike racks
This case exemplifies how misreporting inflates cycling budgets, by including projects and ancillaries that do not relate to the nomenclature of the CI under which they are reported.
Contrarily, the development of cycle paths in the Croatian town of Čepin, categorised under CI 043, proves the existence of significant cycling investment under this CI. The project had a total budget of €1,220,314.24 and an EU contribution of €837,852.03, contributing to the development of 6.3km of cycle paths.
What needs to change
- CI nomenclature should be refined to better represent projects, and projects should be more systematically reported under the CI that best represents them.
- Rules should be established to improve reporting by:
- Standardizing formatting of project titles and descriptions.
- Establishing official definitions of key terms to avoid their misuse.
- Avoiding overlapping CI reporting unless budgets are disaggregated.
- Systematically including geodata for verifiable factchecking.
- Enhance transparency: Address mistranslations, incomplete data, and vague terminology, which can hinder the clarity of project reporting. As elaborated in the report, these are suggestions that apply differently to EU institutions, member states and managing authorities.
Download the report now.
