The present paper estimates and decomposes the employment effect of innovation by R&D intensity levels. Our micro-econometric analysis is based on a large international panel data set from the EU Industrial R&D Investment Scoreboard. Employing flexible semi-parametric methods - the generalised propensity score - allows us to recover the full functional relationship between the R&D investment and firm employment, and to address important econometric issues, which is not possible in the standard estimation approach used in the previous literature. Our results suggest that modest innovators do not create and may even destruct jobs by raising their R&D expenditures. Most of the jobs in the economy are created by innovation followers: increasing innovation by 1% may increase employment up to 0.7%. The job creation effect of innovation reaches its peak when the R&D intensity is around 100% of the total capital expenditure, after which the positive employment effect declines and becomes statistically insignificant. Innovation leaders do not create jobs by further increasing their R&D expenditures, which are already very high.