Clusters: locations, ecosystems, and opportunity
Last newsletter I shared my wonder in Tokyo’s ongoing evolution. The pandemic clearly has not stopped the city’s development.
While in Tokyo I was delighted to give an in-person talk on industry clusters, co-organized by the Japan Society of Northern California (JSNC) ; and GTIE (Greater Tokyo Innovation Ecosystem); at Shibuya QWS (pronounced “cues”). Many thanks to Prof Toru Asahi of Waseda University; Satoru Watanabe and the Japan Society Tokyo team (hat-tip: Yosuke Tanaka); and Shibuya QWS for hosting.

We had a great little cloudburst while I was en route. This let out some of the tension that had been building in the atmosphere. The temperature mercifully dropped a bit.

As JSNC it was our first in-person Tokyo event since….January 2020. It had been a while. It was wonderful to see old friends and colleagues.
The talk’s title translates to: Learning from Silicon Valley, Shenzhen and Pittsburgh – the secrets to industry cluster success!
Hoshikawa-san with Shibuya QWS kicked things off.

Shibuya itself is a sub-cluster within the Tokyo megalopolis – there is a concentration of what telcos would call upper-layer or edge companies – software and game developers and Internet companies like DeNA. LINE (NHN) used to be headquartered there. It’s also the Bit Valley (in retrospect, a not-terrible pun on Shibuya and the Internet) of the dot-com boomlet that happened in Tokyo, and thus, where I did some of my interviewing for my master’s thesis comparing new venture formation in Japan and Silicon Valley many many moons ago. Then, I followed a framework developed from AnnaLee Saxenian’s Regional Advantage. This will be familiar to alums of Clusters class last fall.

Still, giving a talk on industry clusters, in the mega-city, was cause for some meta-reflection. Tokyo (like Seoul, or London) is the political, financial, economic and entertainment center for the country. From a US perspective, it is Washington DC; New York; Los Angeles; and aspects of Silicon Valley combined. Thus, siting a company there, or trying to hire talent there, is the intuitive decision, and the decision for a company is which area of the city, rather than which city. As, say, an international company entering Japan, choosing to site elsewhere may make sense – customer proximity, for example, may dictate setting up outside of Tokyo. But that decision would likely come having also examined Tokyo as an option. Rephrased, not-Tokyo is a conscious choice, made weighing the potential tradeoffs in not being in the mega-city.
This is in contrast to Silicon Valley, or Shenzhen, or Pittsburgh, which can defined not only by their attributes, but what they are *not*, and what they are far from. Silicon Valley is three time zones and a large cultural gulf away from Washington DC. Shenzhen, when it was first designated as a Special Economic Zone, was distant from Beijing and purposefully close to another hub, Hong Kong.
Indeed, the very premise of Saxenian’s Regional Advantage is that Boston Route 128 shares many of the attributes that one would measure when looking at Silicon Valley – universities, students, natural population inflows, transportation infrastructure – and is closer to Washington DC and sources of research funding….so shouldn’t that be advantageous? And, for that matter, some of the original work on the then-Arpanet happened in the Boston area! (Lincoln Lab shows up on early Arpanet maps.) Why, then, would Silicon Valley would ultimately outgrow Boston Route 128 as a venture hub, particularly in the Internet era?

Saxenian highlights culture, and openness, and interpersonal networks and that message resonates as much now as when I first read it.
In my teaching I use a framework expanded relative to Saxenian’s above or Porter’s Diamond framework. Elements of both are included. This is with the goal of getting to a practical, tactical level, from a few potential perspectives: a company thinking of setting up shop or looking to hire; a student thinking about where she might want to live after graduation; the organizer trying to solve issues of common cause (say, drought, or traffic, or nurturing talent).
This is also with an eye to facilitating comparisons between clusters.

I was asked about Pittsburgh during Q&A. The questioner noted that as an example it seemed….different than Silicon Valley or Shenzhen. Yes, precisely! That was intentional.
Pittsburgh is an incredibly relevant comparable for cities dealing with the challenges of de-industrialization and depopulation. If one industry has ebbed, what will be the next to flow? Pittsburgh has excellent universities (Pitt, Carnegie Mellon, Lehigh, among others) and hospitals. It produces tens of thousands of college graduates each year. Its Oakland Square District has a dense concentration of companies and institutions. It has the three rivers that led to the former stadium’s name. It also has 300,000 residents in a city that used to house 700,000, and with that comes a lot of challenges – how to handle empty houses or lots; where to invest and where to disinvest; how to create connectivity tissue between areas.
Indeed, this is why transitions and turnarounds is one of the themes of Clusters class, which starts next week! I’m very much looking forward to version 2.0.