July 17, 2003

Location, Location, Location

One of the most bizarre (from my perspective at least) aspects of the late-1990s high-tech boom in America was its extraordinary concentration in Silicon Valley. With technologies that for the first time allowed a sufficient density of communication that you could do your work and still be well-connected and well-plugged in anywhere with a power socket and a phone line, the premium to being within driving distance or bicycling distance of the Sand Hill Road offramp to Interstate 280 rose to extraordinary and amazing levels. Why?

And will things be different in the "next time" as the next wave of high-tech development accelerates?

: CONSIDER three snapshots of the IT industry today. First, look at the centre of the business at the turn of the century--Silicon Valley. Parts of this high-tech region to the south of San Francisco, which were buzzing with activity only three years ago, are coming to resemble an industrial wasteland populated by squirrels, racoons and "for lease" signs. In the city of Santa Clara, where almost half of all office space is vacant, rents run at around $1 per square foot, compared with $6.50 in late 2000. Some larger tenants in the city are even sub-letting space for nothing, just so they can share the service costs.

Then move north to Redmond, near Seattle, the home of Microsoft. The giant software firm appears to be defying gravity--because it controls two monopolies, Windows and Office, each generating some $10 billion in annual revenues, with operating margins of more than 80%. But the company seems to be running out of bright ideas about where to invest its money. Microsoft has almost $50 billion of cash in the bank. In January, it announced plans to pay a first-ever dividend of $870m this year, a figure that many now expect to be raised, with some talking of amounts as big as $10 billion.

For a third snapshot, turn to Electronics City, near Bangalore in India (above), home of the development centre of Wipro Technologies, an Indian IT services giant. Some 6,500 people work there, about half the firm's total staff, and new buildings are popping up all over the place. By 2005, the campus is scheduled to have three times as many employees as it does today...

Posted by DeLong at July 17, 2003 12:41 PM | TrackBack


I suspect that Silicon Valley is concentrated in Silicon Valley because worker skills are the bottleneck for most projects, and people's skills in the field are difficult to identify and verify. The industry depends upon social networks of workers identifying other workers who have the skills necessary to solve the problems relevant to a specific project. Having a cluster of workers in the same area makes it easier to both socially identify and locally hire needed workers.

Posted by: Jonathan Korman on July 17, 2003 03:04 PM

"The premium to being within driving distance or bicycling distance of the Sand Hill Road offramp to Interstate 280 rose to extraordinary and amazing levels. Why?"

I think the question you raise answers itself - because that's where Sand Hill Road was!

I really do believe that the reason for this concentration on Silicon Valley had everything to do with VCs, rather than the market for talent, as there was certainly plenty of talent to be found in Boston and New York, but they never reached the stratospheric heights reached in the Valley.

Drawing on my own experience, if you wanted to get funding in those years, either you were physically located in the Valley or you might as well not have existed. Most of the prominent VC firms were stating at the time that they weren't interested in funding startups that weren't close to where they were, and that even if they did fund such firms, they'd have to move close by, so the VCs could more easily watch over them.

Nor was the madness confined to the venture capital firms. Many an angel investor seemed to believe at the time that a firm located in the Valley was automatically more of a sure thing than one that wasn't, as if the Bay Area had some sort of monopoly on business acumen or technological innovation.

Posted by: Abiola Lapite on July 17, 2003 03:32 PM

Aye. During the recent boom, there was definitely geographic superstition in the investor community.

Posted by: Jonathan Korman on July 17, 2003 04:02 PM

VC is only one of the many resources that the SV companies share. New companies often spin off of ideas that people have while working for another company. They leave and start their own in their garage or wherever, but why relocate. Then there are the shared resources that develop that service the companies. If you are working with a chip manufacturer, it is more convenient and there is a higher level of interaction in person or on a visit. Info is shared in person or working together that just isn't found in a book or a phone call. Oh, and have you ever tried to get tech support on the phone at one of those companies? Not having to listen to the elevator music alone is reason to locate nearby.

This is the same reason that students go to the universitiy rather than sit at home and get their degree online or from a book. The interactions are at least as important as the information.

Posted by: bakho on July 17, 2003 07:02 PM

I've pondered this question a lot over the last eight years since I left SV. Nearly everything measurable is also available in similar quantities in the large metropolitan areas, especially Boston, NY, DC, and LA.

It's certainly not quality of life.

I keep returning to an almost irrational conclusion: something about UC Berkeley, along with that gauche college to the south that runs a shopping mall on the premises, partly offset then by a sculpture garden nearly always empty at night filled with casts of Rodin masterpieces. (Well I could go on about the Gates of Hell lit up viewed solitarily at night, but...)

There is a rootlessness there that the graduates of these two places seem to embody that fuels the fires of creative destruction. The culture there accepts aids and abets change, and innovation proceeds apace. Add in ample VC money, and you get boom and bust. And Moore's Law.

Posted by: Russell L. Carter on July 17, 2003 08:41 PM


The problem with this statement is that Silicon Valley had no monopoly on talent or manufacturing in the late 1990s. Practically no firms other than Intel, AMD and IBM do their own chip manufacturing nowadays, and Boston has plenty of chip designers, thanks to MIT and all the other universities in the area.

There is also the nature of the IT boom to consider. Most of the really exciting stuff being done was on the software side, not in hardware, so the whole manufacturing issue is for the most part irrelevant. The existence of Linux and the larger Open Source movement demonstrates that proximity is hardly necessary for innovative software development to occur. Neither Mosaic nor Napster were born in the Valley, afterall. Interacting is also hard to do when you're spending half your life choking in smog as you commute from your out-of-the-way dormitory, which was the common plight of many of those I knew who were working in the Valley in the late 1990s.

Add all these things up, and my argument still holds - only the pull of VC funding could have forced people to keep moving to the Valley despite the outrageous overpricing that was going on there. It certainly wasn't rational.

Posted by: Abiola Lapite on July 17, 2003 08:55 PM

As far as I know, UC Berkeley is famous for two things: LSD and Unix. Who can say if one inspired the other? It certainly explains why Unix seems to have been written by someone on a bad trip.

Posted by: Larry Lurex on July 18, 2003 07:02 AM

Absence of a system of vertical trust in development, and the need to use other emerging technologies to get to market at a lower licensing cost encourage proximity.

Open source is the mechanism which creates vertical trust, and new, inexpensive, technologies to leverage.

Posted by: Stirling Newberry on July 18, 2003 07:21 AM

Brad writes "One of the most bizarre (from my perspective at least) aspects of the late-1990s high-tech boom in America was its extraordinary concentration in Silicon Valley."

I think what's missing here is any attention to the fact that there can be local concentrations of something without that something being particularly concentrated. You can have a soap spot on your dark sweater when it comes out of the wash, but very little of the soap is on your sweater.

OK, there were, and are, a few dozen hi-tech firms around Sand Hill Road. The same is true of Queen Street in Toronto (Commodore/Amiga started out its life as Queen Street Typewriter Repair) and dozens more in Brampton, in Markham and God help us in the government-encouraged Ottawa area. In the States there are huge globs of hi-tech in Seattle, in austin, Texas, up and down the Hudson River Valley, and again in all the places the government has force fed it, which is pretty much one dab per Congressional District.

Where's the concentration?

Posted by: David Lloyd-Jones on July 18, 2003 09:55 AM

" With technologies that for the first time allowed a sufficient density of communication that you could do your work and still be well-connected and well-plugged in anywhere with a power socket and a phone line,..."

This statement is really only true half the time. In the tech industry, its often the case where multiple people need to be physically in the same location, to resolve an issue. Sometimes its a limitation on expensive hardware. Often, its the inability to accurately communicate without the physical presence. Technical conversations, seem to take twice as long over the phone. Even longer over email.

The second reason is the availability of recruiting new talent. Boston's 495 has (had) a similar setting. Motorola, Digital, etc. Being in the same state as the tech schools aids in recruiting. Its also easier to draw in interns and grad students for the summer.

Posted by: james on July 18, 2003 10:37 AM

Although this book is from the mid-90s, it's still relevant to this discussion: Annalee Saxenian's Regional Advantage, "culture and competition in Silicion Valley and Route 128" in which the author discusses the importance of a decentralized and cooperative industrial system in Silicon Valley (which she contrasts with the Route 128 area around Boston).

Posted by: Eszter on July 18, 2003 10:41 AM

"I've pondered this question a lot over the last eight years since I left SV. Nearly everything measurable is also available in similar quantities in the large metropolitan areas, especially Boston, NY, DC, and LA."

I think it comes down to Shockley having grown up in Palo Alto.

There are a couple of other reasons, in the form of institutions, UC Berkeley, that Junior University in the peninsular, Livermore, LBL, SRI, Xerox PARC, IBM Alamaden, HP. Oh, and for Biotech, UCSF (hence Genentech & Chiron, & hence the other biotechs in Bay Area).

Also, you have a huge amount of social capital in networks; informal & professional societies, alumni associations, and a large "gift economy" of advice & contacts. I was part of an abortive biotech startup out of UCSF & UCB B-school; I was amazed how much time you can get from senior people, based I presume that people want to be nice to you on the way up, so you can later return the favor.

"It's certainly not quality of life."

Speak for yourself. I saw a live theatre two weeks ago where a napkin-and-teacup puppet did a parody of "Crouching Tiger, Hidden Dragon". Can you see art like that in the Research Triangle?

Posted by: Tom on July 18, 2003 11:21 AM

I would rephrase the question. Have people in Silicon Valley run out of ideas, infrastructure (financial or otherwise), or people.

Apple seems to continue to innovate. After trying there new operating system I can't help but think how much better it is than windows. And I have three computers in my office. A windows machine, and old g3, and a laptop with 10x on it. I find myself using the lap top for the sheer pleasure of working with the operating system. Ideas are still there.

People for the most part are. Did they all go to Redmond.

On the other hand I can't help but think the Microsoft anti-trust case sucked the life out of tech generally in this country. Microsoft, for all its money , is not innovating. And it certainly has the large amounts of cash. Similarly, i think tech firms generally in the US are doin badly.

Posted by: Lawrence on July 18, 2003 02:14 PM

Well, the biggest factor that's not being addressed so far in these comments is that the tech boom required relatively skilled workers and so most (but not all) of the concerntrations of tech industries occured in cities with very high educational density coupled with some existing high-tech (since there's an overlap in the workforce).

I think a good example of this is how it played out here in Texas. Austin was one of the four or five high-tech centers of prominence in the US after SV. Why? Why wasn't it Dallas/Ft. Worth, or Houston, or even San Antonio? I mean, these cities, including Austin, taken together form a north-south corridor that's one of the biggest urban corridors in the US. I think the fact that Austin got the biggest concentration of tech was these two factors: preexisting high tech and education. Dallas has always been a finance and services concentration, Houston industry (and both oil, but not Austin). Dallas was/is a center of communications tech, but really not much else in terms of high tech. I don't know about San Antonio, because even though it's well over a million people, I don't know what the heck they do there. (It's one of those invisible large cities, it seems.) But Austin has had a high tech presence for about twenty-five years, or so, distributed between manufacturing (like a few chipmakers with fabs), other hardware (for example, IBM--this is where the HQ of the rs6000s and AIX is), and services (again, IBM, others). Then the biggest thing is one of the very biggest public universities that is also well-regarded. UT Austin was enormously large relative to the Austin of the 80s, and still dominates much of the character of the city. I don't know how much being the state capitol helped, but it probably did, too. But Dallas has always seemed to me to be higher education "poor" relative to other cities its size, and this is true of Houston, too. And it's particularly true of San Antonio. Now look at Austin, the Bay Area, Boston, the Research Triangle of North Carolina: education centers.

So, Austin was one of the top four or five fastest growing metro areas in the US in the 90s, and much of the job growth was in the tech sector. It's why I moved here (well, my best friend is from the region and he moved here first).

What's interesting is that even though the US labor force is very mobile, and probably the tech workers were moreso, and even though Austin companies recruited nationwide, the tech labor shortage was noticably worse here than it was elsewhere (excepting the other tech concentrations). So employers were paying huge premiums for labor that high tech employers in, say, Albuquerque (where I'm from, which has a mildly prominent tech sector) weren't. That's inefficient.

Now the reverse is true--the economy here isn't as bad as it is in SV, but it's worse than average. There's a huge surplus of skilled tech workers and no jobs whatsoever. Now there's the reverse inefficiency. I don't doubt that there are some areas of the US where there's still a tech labor shortage.

I'd like to see some (comprehensible to a layperson like myself) academic economic discussion about the forces at work in this sort of situation. Even though the economic sector involved should have not needed to be so localized, and the labor force often didn't even need to be local (though it almost always was--how come?), and even though the US labor force is notably mobile....we still see in how this played out that there's lots of inefficiency. How come? Is this improving?

Posted by: Keith M Ellis on July 18, 2003 03:03 PM

"Speak for yourself. I saw a live theatre two weeks ago where a napkin-and-teacup puppet did a parody of "Crouching Tiger, Hidden Dragon". Can you see art like that in the Research Triangle?"

I saw that done by the San Diego Chicken (now the Famous Chicken, I guess) at a Durham Bulls game a couple weeks ago. ;-)

Seriously, though...most accounts that I read of why the Research Triangle isn't comparable to Silicon Valley is that there are far fewer entrepreneurs in the Research Triangle.

Plenty of people work at Glaxo-Smith-Kline (Beacham-Burroughs-Wellcome ;-)). Plenty of people work at IBM's many Research Triangle Park (RTP) facilities. Plenty of people work at SAS in Cary (not actually in RTP, but in the Triangle area).

But people stay at those places...they don't venture out to form their own companies.

One problem the Research Triangle doesn't have: the Research Triangle area is about as nice a place to live as one could pick. Moderate housing prices. Good schools, including the 3 universities that form the Triangle. Very pleasant climate (though it would be h@ll without air conditioning). Commuting times aren't outrageous. One can travel pretty much anywhere along the East Coast quickly, at a reasonable price, and without jet lag.

Posted by: Mark Bahner on July 21, 2003 10:01 AM

Michael Porter's work on regional specialization is worth considering here. An interacting complex of local supplies, demands, and sophistications create both a concentration and speed of informal information and formal execution that gives a specialized regional economy like Silicon Valley a renewable edge. Specialized services (design, equipment, talent, marketing, finance) are not only available, but are able to provide a large and lucrative enough market for workers who want to hone specific skills makes possible not only the availablity of unique local abilities, but allows unusual talents to develop that could not be supported in a region which did not have such a high concentration of specialized activity.
Sophisticated customer demand also plays a critical role. Much of Silicon Valley's success story lies outside the large well known companies, in small specialized service firms and temporary ventures that draw together unique talent to solve particular technical problems. This is to some degree a phenomenon of "internal customers," in which people inside an industry become aware of needs that create entrepreneurial opportunities unknown to the larger market. The combination of many strengths available together along with a culture of flexible application of those strengths to insider industry problems is a powerful advantage, and given the fact that networked computing was creating real economic efficiencies for such a large portion of companies in the general economy, Silicon Valley was really in the right situation at the right time.
There seems no real reason, however, that such a regional advantage cannot be duplicated elsewhere, especially in places like Bangalore, India, where such a huge cost advantage exists. Given the extrordinary access India now has to both Silicon Valley and engineering education all over the US, the specialized abilities formerly available only in the US no longer constitute an inassailable comparative advantage. Whether there will remain advantages which can overcome the tendency to extrordinary increases in the cost of operations which made Silicon Valley such a difficult environment during the late 1990s is a key question. The landholding community clearly overplayed their hand, and this one factor seems quite out of step with the cooperative nature of the other factors in SV's economy. Silicon Valley is a valuable resource, but still may find itself becoming the Detroit of the future, as international competition and national economic development policies compete for technical industries.
In the late 1980's, Asian electronic industries missed the digital boat, having focused on analog electronics, and this misstep left Silicon Valley in a lucky position of advantage during the 1990's, an advantage which has ended. Whether this kind of discontinuity was an historical fluke, or whether it was an example of just the kind of advantage of insider sophistication that gives a place like Silicon Valley a continual advantage, may be the kind of question that determines whether SV is like New York or just a new Detroit. In terms of culture, architecture, art, and the general evidence of rising beyond a mere industrial workzone, Silicon Valley has yet to provide much value to its citizens. Without its industrial domination, would the suburbs and factories South of the bay even retain the interest of those who live there? Unlike the great cities of the world, the moguls of Silicon Valley haven't shown much pride or interest in developing the kind of public space that holds intrinsic merit. This may come in time, but whether Silicon Valley is more akin to Liverpool or London remains to be seen.

Posted by: B Price on July 29, 2003 12:13 PM
Post a comment