Did R&D Misallocation Contribute to Slower Growth?
Abstract
This paper provides evidence that rising frictions and their impact on the allocation of R&D resources have contributed to the slowdown in US productivity growth in recent decades. I develop a growth accounting framework allowing for rich firm heterogeneity in R&D productivity, exposure to frictions, as captured by an R&D wedge, and the rate at which private value created from innovation translates into growth. The model growth rate permits a closed-form decomposition into the frictionless level and an adjustment factor capturing the impact of frictions. I propose a methodology to measure the model primitives for a sample of US-listed firms from 1975 to 2014. Frictions can be measured using the R&D return, i.e., the ratio of value created from R&D to its cost, which I measure as the ratio of patent valuations to R&D expenditure over a 5-year window. I document large and persistent differences therein, which suggests significant frictions through the lens of the model. The evidence suggests financial frictions, adjustment costs, and monopsony power over inventors as potential drivers of R&D return dispersion. Combining model and data, I estimate that frictions reduce economic growth by 18% on average and that their rise can account for 11% slower growth in 2000–14 or 30% of the overall observed slowdown. These findings are robust to a large set of alternative specifications and measurement error adjustments.
