Enrico Moretti (Berkeley)

“The Effect of  High Tech Clusters on the Productivity of Top Inventors”

Abstract: The high tech sector is geographically  concentrated in a small number of expensive cities. The degree of geographical clustering by research field is very high, with the top ten cities in  “Computer Science”, “Semiconductors” and “Biology and Chemistry”,  accounting for 70%, 79% and 59% of inventors in their field, respectively.  These shares have been growing since  1971, indicating increased agglomeration. Why do high tech inventors tend to locate near other inventors in the same field, despite the higher costs? I use longitudinal data on top inventors based on the universe of US patents 1971 – 2007 to quantify the productivity advantages of Silicon-Valley style high tech clusters, and their implications for the overall production of patents in the US. I relate the number of patents produced by an inventor in a year—a measure of inventor-specific productivity—to the size of the local cluster, defined as a city times research field. I first study the experience of Rochester NY, whose high tech cluster declined due to the demise of its main employer, Kodak.  At its peak,  Kodak accounted for most inventors in Rochester, but due to the diffusion of digital photography, its employment collapsed after 1996, resulting in a 49.2% decline in Rochester high tech cluster.  I test whether the change in cluster size affected the productivity of inventors outside Kodak and the Photography sector.  I find that between 1996 and 2007 the productivity of non-Kodak inventors in Rochester declined by 20.6%  relative to inventors in other cities in the same field, conditional on inventor fixed effects. In the second part of the paper, I turn to estimates based on all the data in the sample. I find that when a inventor moves to a larger cluster, defined as a cluster with more inventors in a given field, she experiences significant increases in the number of patents produced and the number of subsequent citations.    Conditional on inventor, firm fixed  and city times year effects, the elasticity of number of patents produced with respect to cluster size is 0.0662. The productivity increase follows the move, and there is no evidence of pre-trends. Estimates appear robust to different assumptions on selection.
Instrumental variable estimates based on the geographical structure of firms with laboratories in multiple cities yield an elasticity larger  than the OLS elasticity.
In the final part of the paper,  I use the estimated elasticity of productivity with respect to cluster size  to quantify the aggregate effects of geographical agglomeration  on the overall production of patents in the US. I find  macro-economic benefits of clustering for the US as a whole. In a counterfactual scenario where the quality of U.S. inventors  is held constant and their geographical location is changed so that all cities have the same number of inventors in each field, inventor productivity would increase in small clusters and decline in large clusters.  On net, the overall number of patents produced in US in a year would be 11.07% smaller.