EU Innovation Fund Project-Level Dataset for Post-Signature Termination Analysis
Description
This repository contains the project-level analytical dataset and reproducibility files supporting the manuscript “From Award to Delivery: Project-Level Correlates of Termination in the EU Innovation Fund. The dataset is constructed from publicly available European Commission and European Climate, Infrastructure and Environment Executive Agency (CINEA) sources. It covers regular-grant projects from the 2020–2023 EU Innovation Fund calls included in the empirical analysis. The analytical sample contains 207 funded projects, of which 45 were classified as post-signature terminations at the data-freeze date. The repository includes the cleaned analytical dataset, a data dictionary, the frozen Python analysis code, and reproducibility documentation. Variables cover awarded Innovation Fund grant, expected greenhouse-gas avoidance, project status, project age at the data-freeze date, technology group, project scale, call cohort, and derived analytical variables. The Innovation Fund Support Intensity (IFSI) variable is an analyst-constructed measure defined as awarded Innovation Fund grant divided by expected absolute GHG avoidance. It is not an official European Commission selection or cost-efficiency metric. No confidential data are included. Source provenance is documented in the accompanying files.
Files
Steps to reproduce
1- Download all files in the repository. 2- Place IF_full_clean.csv in the location specified in README_reproducibility.md, or adjust the input path in analysis.py if necessary. 3- Run analysis.py using Python 3.13 or a compatible Python environment with the package versions documented in the README. 4- The script reproduces the descriptive statistics, Firth bias-reduced logistic regression specifications, the IFSI specification, restriction tests, and robustness checks reported in the manuscript. 5- No imputation is performed. Sample construction, variable definitions, exclusions, and source provenance are documented in data_dictionary.md and README_reproducibility.md.
Institutions
- Universidad de BurgosCastille and León, Burgos