Open Access

Maria del Sorbo, Carina Faber, Marco Grazzi, Francesco Matteucci, Miriam Ruß

• A novel analysis of the European Innovation Council (EIC) Accelerator pilot is presented, marking the first extensive examination of its selection process and the impact of its funding on deep tech ventures, in comparison to its predecessor, the SME Instrument. Utilizing applicant data from both programs, the study assesses the EIC's effectiveness in targeting firms that align with its objectives of driving breakthrough innovation. The research reveals that the EIC Accelerator pilot attracts younger and smaller firms, in comparison to its predecessor. A significantly higher proportion of applicants are high tech and medium hightech, indicating a strategic shift towards supporting cutting-edge technologies. Despite this shift, the analysisA novel analysis of the European Innovation Council (EIC) Accelerator pilot is presented, marking the first extensive examination of its selection process and the impact of its funding on deep tech ventures, in comparison to its predecessor, the SME Instrument. Utilizing applicant data from both programs, the study assesses the EIC's effectiveness in targeting firms that align with its objectives of driving breakthrough innovation. The research reveals that the EIC Accelerator pilot attracts younger and smaller firms, in comparison to its predecessor. A significantly higher proportion of applicants are high tech and medium hightech, indicating a strategic shift towards supporting cutting-edge technologies. Despite this shift, the analysis of funding determinants demonstrates a consistent pattern across both programs, emphasizing the influence of firm size, age, and patent portfolio. Further, a regression discontinuity design analysis is used to estimate the impact of funding during the EIC accelerator pilot on firm-level outcomes, such as patenting, revenue, or employment growth. However, the very recent launch of the program shrinks both the observations and the ex-post window, and due to large standard errors the point estimates are not significant at conventional levels.…