| Abstract: |
This paper quantifies the determinants of R&D investment heterogeneity and its
implications for growth. We estimate a Schumpeterian growth model with
heterogeneous firms, à la Lentz and Mortensen (2008), in which firms differ
with respect to innovation efficiency. Using observations on size,
productivity, and R&D expenditures from a panel of Norwegian manufacturing
firms we find that the model has a good fit to the data. In particular, it
fits the distribution of R&D investment (mean, dispersion and skewness) as
well as the negative correlation between research intensity and size.
Moreover, the model generates firm-level investment responses to R&D subsidies
that are in line with micro evidence from a natural experiment. The model
estimates imply that a large part of aggregate productivity growth (72
percent) is the result of the market directing R&D resources to the more
innovative firms. Finally, we study the link between firm heterogeneity and
R&D subsidies, and show that the growth effects of subsidies depend crucially
on how the policy influences the equilibrium distribution of firms. We address
these questions by estimating an equilibrium model of firm-level innovation
and growth. We adopt the micro-macro framework growth in Klette and Kortum
(2004). Using observations on size, productivity and R&D expenditures from a
panel of Norwegian manufacturing firms, we estimate an extended version of the
Klette-Kortum model, developed in Lentz and Mortensen (2008). The model
estimates imply that a large part of aggregate productivity growth (72
percent) is the result of the market directing R&D resources to the more
innovative firms. Using the estimated model we also show that the growth
effects of R&D subsidies depend crucially on how the policy influences the
equilibrium distribution of firms. |