Knowledge Spillovers and Local Industry Growth: A Patent Citation Approach

Abstract

We create unique patent-based measures of Marshall–Arrow–Romer (MAR) and Jacobs knowledge spillovers using patent citations data and use them to test the Glaeser et al. model of local industry employment growth on three emerging technology categories, namely, computing and communications, drugs and medical, and electrics and electronics. We test growth in 45 U.S. metropolitan statistical areas (MSAs) and consolidated metropolitan statistical areas (CMSAs) for eight two-digit industries over the period 1994 to 2000. We find strong evidence for MAR spillovers from specialization, but little for Jacobs spillovers from diversity. Our results suggest that regional specialization, but not diversity, boosts local industry employment growth in these knowledge fields. JEL classifications: J21, L16.