Our work applies the model developed by Acemoglu et al. (2018), henceforth, AAABK, for assessing the growth and welfare implications of different types of innovation policies. Central to the AAABK model is the ratio of high-productivity and low-productivity firms in total output and how different policy measures affect this relationship.

We employ the AAABK framework in order to build a macroeconomic model of the innovative business sector in Finland and fit it to the company-level micro-data on Finnish companies from 2000 to 2016. Acemoglu et al. (2018) employed US data from the 1980s and 1990s. We complement their work by estimating the AAABK model for more recent years: 2000–2016. Our results add to the literature by providing evidence on the aggregate effectiveness of innovation policies in this more recent period of slower economic growth.

Our empirical findings yield, by and large, similar qualitative conclusions on the effects of public policies on economic growth and welfare to those reported in the original work using the US data. Generally, increasing R&D subsidies would be a recommendable policy. The welfare impacts of R&D subsidies are highest when they accelerate the re-allocation of R&D workers to companies with high R&D productivity. The most effective innovation policy targets R&D subsidies to companies with the highest innovation capacity (i.e., in these companies, R&D employees generate the highest increase in a firm’s productivity). If subsidies are allocated to companies with low innovation capacity or to low-productivity companies that are close to exiting the market, there will be less innovation and slower economic growth.