R&D Agglomeration, Spillovers, and Recombination: Mechanisms and Implications for Productivity and Growth

Paper Session

Saturday, Jan. 7, 2017 8:00 AM – 10:00 AM

Swissotel Chicago, Zurich A
Hosted By: American Economic Association
  • Chair: Bruce A. Weinberg, Ohio State University, IZA, and NBER

The Mechanics of Endogenous Innovation and Growth: Evidence From Historical United States Patents

Ufuk Akcigit
,
University of Chicago and NBER
William R. Kerr
,
Harvard University, NBER, and Bank of Finland
Tom Nicholas
,
Harvard University

Abstract

How does technological progress occur? Is the nature of innovation stable over time? We shed new light on these questions through a mixture of empirics and theory. We begin with an empirical analysis of patents granted by the United States Patent and Trademark Office. This analysis reveals several striking facts that emphasize the increasing importance of novel combinations of technologies for U.S. patents, compared to either new technology development or the reuse/refinement of older technology combinations, and the localized nature of these recombinations. We build an endogenous growth model that can match these facts and illustrate the underlying mechanics of the technological development process.

R&D Spillovers and Scientist and Engineer Mobility

Erling Barth
,
Institute for Social Research and NBER
James C. Davis
,
U.S. Census Bureau
Richard B. Freeman
,
Harvard University
Gerald R. Marschke
,
State University of New York-Albany and NBER
Andrew Wang
,
Harvard University and NBER

Abstract

Using Census data that link U.S. workers to their workplaces we estimate how much past employment in R&D-performing firms raises scientists' and engineers' (S&E) wages and thus their marginal product at new employers. We find that S&E workers who arrive at a new firm that is R&D-active having been exposed to R&D in previous employment are more productive. We then use Census data to estimate establishment-level production functions showing that both the establishment’s own R&D and R&D activities by other firms that are geographically close and in the same industry have positive effects on establishment output. This evidence is consistent with previous studies. We then add to the production function a measure of external R&D brought to the establishment via newly hired high-wage workers. The coefficient estimate on incoming human capital-embodied R&D is both statistically and economically significant, but does not substantially reduce the coefficient estimates on external R&D. While this evidence is highly preliminary, it suggests that while worker mobility plays a role, it is not the dominant mechanism by which R&D spillovers among firms occur.

An R&D Design to Assess the Causal Impact of Tax Policy on R&D, Innovation and Spillovers

Antoine Dechezleprêtre
,
London School of Economics and Political Science
Elias Einio
,
VATT Institute for Economic Research and London School of Economics
Ralf Martin
,
London School of Economics and Political Science
Kieu-Trang Nguyen
,
London School of Economics and Political Science
John Van Reenen
,
Massachusetts Institute of Technology

Abstract

We present the first evidence showing causal impact of research and development (R&D) tax incentives on innovation outcomes. We exploit a change in the asset-based size thresholds for eligibility for R&D tax subsidies and implement a Regression Discontinuity Design using administrative tax data on the population of UK firms. There are statistically and economically significant effects of the tax change on both R&D and patenting, with no evidence of a decline in the quality of innovation. R&D tax price elasticities are large at about 2.6, probably because the treated group is from a sub-population subject to financial constraints. The policy generates significant innovation spillovers to other “technologically close” firms. We estimate that over 2006-11 business R&D would be around 10% lower in the absence of the tax relief scheme.

The Location and Agglomeration of Industrial R&D

Sifan Zhou
,
Harvard University

Abstract

Using the Survey of Industrial Research and Development (SIRD), the Longitudinal Business Database (LBD) and the USPTO patent data, this paper conducts the first systematic and longitudinal mapping of industrial R&D activities in the U.S. and describes the degree of industrial R&D agglomeration over the period between 1978 and 2011. I then study the determinants behind such agglomeration of industrial R&D by studying individual firms’ location decisions on where to open new R&D establishments. The first set of determinants are the locations of a firm’s pre-existing establishments. On the one hand, many R&D activities, particularly those aimed at process improvement, are not easily separated from manufacturing. This gives firms incentives to collocate R&D labs with their own manufacturing plants. On the other hand, R&D demands highly skilled, highly paid workers and manufacturing demands low-skilled, low-paid workers. This tends to pull R&D activities away from manufacturing. The second set of determinants relates to the knowledge spillovers across firms within industry. Locating close to competitors’ R&D facilitates the absorption of new technologies, skills and ideas from these competitors, whereas locating R&D facilities remotely limits one’s own knowledge from “spilling over” to competitors and helps the firm fully internalize the returns to its R&D. The third set of determinants come from outside of the industry, including the proximity to research universities, access to highly educated and highly specialized labor markets, regional IP protection such as the enforceability of non-compete agreements, and taxes on R&D expenditures. I examine how the direction and relative importance of these factors on R&D lab location and agglomeration vary by the type and stage of R&D.
Discussant(s)
Ajay K. Agrawal
,
University of Toronto and NBER
Benjamin F. Jones
,
Northwestern University and NBER
Timothy Simcoe
,
Boston University and NBER
Daniel J. Wilson
,
Federal Reserve Bank of San Francisco
JEL Classifications
  • L1 - Market Structure, Firm Strategy, and Market Performance
  • O3 - Innovation; Research and Development; Technological Change; Intellectual Property Rights