Economic Downturns, Technology Trajectories and the Careers of Scientists
Akinsanmi, Eyiwunmi O.
Reagans, Ray E.
MetadataShow full item record
This paper studies the effect of the burst of the telecom bubble on the trajectory of an emerging technology, and the careers of scientists in that industry. We focus on optoelectronics, a general purpose technology (e.g. Helpman 1998) with applications in energy, biomedical, telecommunications, computing and military. Leveraging USPTO patents, we analyze the relationship between an inventor’s pre-bubble characteristics and his productivity post-burst and thereby the national trend. Past research in technology innovation has used publicly available patent data to measure productivity (Kapoor and Lim, 2007, Fleming 2007), inventor mobility (Song et al, 2003, Rosenkopf and Correidora, 2009, Marx et al 2009), and transfer of knowledge (Almeida and Kogut, 1999, Song et al, 2003, Rosenkopf and Correidora, 2009). In this research we use patent data to estimate inventors’ pre- and post-bubble burst knowledge capital (here, number of patents), pre- and post-bubble burst mobility (i.e. number of assignees), pre- and post-bubble burst technology focus (based on patent classes), and pre-bubble burst career length patenting in optoelectronics (here time from first to last optoelectronic patent, allowing for a period of inactivity based on gaps in one’s patent applications). With respect to technology focus, the dimension on which we differentiate inventors is whether they have pre-bubble patents in “integration” – an emerging optoelectronics technology that facilitates optoelectronics’ application to non-telecommunications applications. We use logit and OLS models to explore potentially interesting relationships between inventors’ pre-burst characteristics (knowledge capital in general or non-integration OE and knowledge capital in integration, career length, and their interactions) and (1) continuing to patent in optoelectronics and (2) productivity post-burst. We estimate simple slopes (Aiken, 1991) using the margins command in Stata 11. Our model finds that on average higher pre-burst knowledge capital and career length predict higher probabilities of continuing to patent in the field; however, the probability that an inventor will continue to patent if he has high pre-burst knowledge capital in integration is higher than if he has pre-burst knowledge capital in general OE. In addition, we find that moves by pre-burst integration inventors who patent in integration post-burst are associated with higher productivity while moves by pre-burst general optoelectronics inventors (without patents in integration) who patent in optoelectronics post-burst are associated with lower productivity. This differentiated role for inventors who have ever patented in integration supports past work suggesting that new technologies can experience disproportionate advancement during economic downturns, and highlights the need for further research into the relationship between sector specific downturns and the diversification of technologies into new market applications.