The making of high-tech spillovers: evidence from early industrial labs

Technology breakthroughs often give rise to unanticipated early leader firms in a nascent industry. Do these early leaders cease to share knowledge in public after they achieve product market leadership? Using newly digitized plant-level data and a unique natural experiment based on pre-Fairchild semiconductor choices, I test the corporate response to a plausibly unanticipated head start in the American late 1950s’ microchip breakthrough. I find: 1) The head start quickly creates early leaders in the nascent microchip industry, marked by persistent influx of patent citations and expansion of product lines. 2) Product differentiation surges: the early leaders persistently announce new products in their leading research fields. 3) Rather than fencing off these research fields, the leaders selectively present applied research as public knowledge in industrial associations, and release increasingly more technical manufacturing bulletins. I provide a simple theory to illustrate the early leaders' trade-off in deciding the optimal scope of public knowledge sharing: on one hand, presenting techniques in public unavoidably benefits rivals; on the other hand, presented knowledge resembles persuasive advertisement that raises perceived product differences for consumers. 4) The selective sharing demands face-to-face contact, likely facilitating agglomeration externalities in early leader locations.