R&D Collaboration and the History of Bell Labs

In this series of posts, I use Jon Gertner’s history of Bell Labs as a jumping-off point to explore 5 aspects of innovation.  This is the second post in this series.

When should an R&D lab pursue external relationships?  With whom, and to what end?  What should those relationships entail?

Bell Labs: Informal Partnerships

In Jon Gertner’s The Idea Factory, we see that Bell Labs was never an isolated institution.  It maintained a number of informal relationships with academia, and even other firms.  Both people and ideas flowed to and from Bell Labs.

Many of Bell Labs’ earliest employees were sourced from the University of Chicago, and in particular were sent there by one professor – Robert Millikan (p.27).  Unsurprisingly, many more came from other great engineering schools, like MIT and Caltech (p.38).  The barriers between academia and Bell Labs were apparently permeable in both directions; numerous scientists resumed careers in academia after stints at Bell Labs, and some even ping-ponged back and forth multiple times.

Bell Labs’ university relationships extended beyond labor churn; ideas flowed freely as well.  Labs’ scientists kept up on the latest science through academic journals, through visitations to top universities, and through a regular Monday colloquium that hosted lectures by outside scholars like Enrico Fermi.  As with people, the exchange of knowledge went both ways: Bell Labs sent out its own academic publication, and its scientists gave their own lectures at various universities (p.41-3).

Less intuitively, Bell Labs also maintained relationships with rival firms, sharing knowledge in much the same way (despite the possibility of losing competitive advantage).  Ian Ross, who worked at Bell Labs on transistors in the 1950s, termed this the “‘precompetitive era,’” when “information was freely exchanged.”  At this time, companies like Texas Instruments and Fairchild Semiconductor were merely startups, and their employees would swing by Bell Labs just to attend seminars and see what was new (p.251-2).  Even though these companies would grow into giants in the space, this earlier period—when the technology was young—appeared to foster more collaboration than competition.

There is certainly a pattern to Bell Labs’ external relationships, but it is so far merely correlative.  Was Bell Labs a great, innovative institution in-part because of these external relationships, or was the effect modest, zero, or even negative?

The Academic Literature

While the literature is inconclusive, there does appear to be strong empirical support for Bell Labs’ relationships being a net positive.  Two streams of literature are most pertinent here, both outlined and described by Agrawal (2001).    

The first stream of related research is on “absorptive capacity,” originating with a highly cited paper by Cohen and Levinthal (1990).  This literature investigates firms’ capacities to make use of the knowledge generated at universities.  Earlier papers posited that this capacity was related to overall investment in R&D, while later research emphasized the importance of university-connectedness in being determinative both of R&D investment and absorptive capacity.  While these results may seem obvious, this was a crucial first step in establishing the baseline importance of university-firm connectedness.

The other major stream of related research illuminates the nature of that connectedness, and focuses on the channels of knowledge transfer between universities and firms.  This set of papers examines the relative importance of many knowledge pathways, including patents, recruiting students, conferences, publications, consulting, and informal conversations.  While there is variation in the relative importance of each by industry and by individual invention, patenting appears to be among the least critical, while those factors associated with ‘open science’ (like publications and conferences) are among the most important.

Agrawal also points to a couple of promising areas for future research.  These include a deeper exploration of variation across industries, as well as the specific mechanisms underlying some of the less-quantifiable knowledge-transfer channels, like consulting and recruiting graduate students.  Given Gertner’s case evidence from Bell Labs, we might add to this wish-list an additional stream of research: how information sharing and exchange changes over time as an industry/technology matures.

Overall, we can say with relative confidence that external relationships have a positive effect for firm R&D efforts, but plenty of work remains in specifying the mechanisms (and, ultimately, being able to supply prescriptive advice for firms and universities). 

More from Kauffman

comments powered by Disqus