Is Silicon Valley Doomed?

There is no shortage of pundits proclaiming the death of Silicon Valley and the greater San Francisco Bay Area (BA).  Certainly, this metropolitan region has enjoyed tremendous economic growth over the past 2-3 decades.  But, given today's anemic job growth, ever consolidating software companies and inevitable loss of jobs, global info-worker wage competition, and of course those ever annoying BA real-estate prices, is there any hope of a real recovery this time?

The San Francisco Bay Area is the fifth largest metropolitan region in the US.  It is also among the most expensive places to live in the country, primarily due to the past ten years during which BA real-estate has appreciated at an unprecedented rate.  Other things are expensive in the BA, too.  Food is expensive.  Gas is expensive.  Movies are expensive.  Even crossing one of our many bridges is getting more expensive seemingly every year.  With all this inflation, one would expect that salaries have also inflated.

Not the case. Since the popping of the dot-com bubble, and the subsequent recession, BA salaries have failed to keep pace with inflation.  In fact, there is a great deal of dispute currently about whether there has actually been real wage deflation.  Certainly, for those workers unfortunate enough to have found themselves on the receiving end of off-shoring and global wage arbitrage, there have been real hardships.

There is a deeper dynamic at work.  Generally referred to as "wage-efficiency-theory", it basically says that workers who manage to keep their jobs through adjustments like we're experiencing now do quite well.  However, workers who lose their jobs, or even move their jobs voluntarily, lose out as they must accept lower wages.  Anecdotally, most in the BA see this happening before their own eyes.  There are no shortage of blog discussions where "the salary for a Java engineer" is hotly debated, with some swearing their companies pay >$150K while others retort that they can't find an equivalent job for anything breaking the $100K point.  They are both right.  This is wage-efficiency-theory at work.  Companies don't react instantly to changes in the labor market, and so wages are reset through a slow process of attrition.

But there has also been growth too.  Unfortunately, a large portion of recent growth has come from real-estate and related industries.  Most economists, and now a widening consensus of experts, agree that such growth is ultimately not sustainable.  As the real-estate bubble itself eases -- or outright pops -- and many of the real-estate related jobs evaporate or cheapen, will there be a new engine of growth to carry the BA into another dawn?

A final thought is that the BA has been here before.  Software is not the Valley's original economic engine.  Before this we had chips (thus the name Silicon Valley).  There was talk of the BA becoming the New Detroit then, but that was ancient history and many billions of dollars of GSP contribution ago.  The question is, can bio-tech, nano-tech, quantum computing, DNA computing, genetic engineering, and even new media (Web 2.0 stuff) coalesce into the basis for another virtuous cycle in the San Francisco Bay Area?

--by Randolph Harrison

Comments

[Peter P]

Perhaps we can compare the Bay Area with the Boston area, which also as top notch universities and capital intrastructure. One advantage of San Francisco is the relative proximity to Asia. What else?

[Peter P]

Also, another question will be: Silicon Valley is doomed for whom?

[randolfe]

Good question. There are always winners and losers. I was thinking much more macro-level. But, it could well be that any BA rebirth will cause a lot of displacement (economic and geographic) while simultaneously fertilizing new ground for new industry.

I would have said that I'd expect very specialized engineers and technologists to take the brunt of resturcturing. But that has largely already occurred due to global outsourcing pressures. Perhaps, in this increasingly global economy, restructuring happens on an ironically longer timescale; maybe a bit smoother. ie., no Flint Michigan type of collapse.

[randolfe]

I definitely consider Boston to be an equivalent geographic nexus capable of virtuous cycles. I'm not a Boston expert by any means so I leave it to others to help fill in the details beyond the obvious. Consider the Bay Area's Asia link, is Boston an equal as a EMEA link?

Perhaps the areas which will take it hardest are those that have developed economies similar to Boston and the Bay Area on the tail of the tech economy success, but which lack one or more key ingredient to be a true innovation nexus. A couple of areas come to mind.

[Peter P]

You are recommending quite a few books on genetic algorithms. Are you currently using this technology actively? I did think of using GA to perform portfolio optimizations but I just do not have the time and patience to build the darn thing. :)

[randolfe]

I used GAs in a network element system I created for SBC. The GAs allow for dynamic, self optimizing search of the network. The network changes over time, and itself is not mathematically optimized, so GAs seemed like a reasonable solution.

All that said, I am still not convinced that GAs were the most economical solution, even though the subsystem has performed well and held up for 3.5 years now in production. GAs are really tough to train, and very hard to find programmers to maintain. Big ugly state machines may not be sexy, but they are usually economical.

[Peter P]

"Big ugly state machines may not be sexy, but they are usually economical."

I guess it is all about Evolution vs Intelligent Design. ;-)

But GA is not bad for a changing system. I like it better than neural networks because it is more intuitive.

[randolfe]

I love neural networks, but have yet to find an excuse good enough to get someone to actually pay me to implement one. Do you have any good references of practical applications people are using outside of big military & academic circles?

[Peter P]

Do you have any good references of practical applications people are using outside of big military & academic circles?

I heard that some are using it for trading. The problem is that the trained model is not human-readable at all. It is just difficult to have faith in something like that.

GA is more flexible. It can be used to generate/evolve rule-based system. It can even be used to train neural networks. :)

[lost_in_silly_valley]

BA has been doomed before. Anecdotal evidence from my recent job search. I was looking at engineering / engineering management positions before landing in something entirely different. Java engineer salaries greater than 150 k? Very rare. 100 -120 k is more like it. Director level engineering positions in large software companies 130 - 160 k.
The median house prices in Cupertino is around 800k, Saratoga 1.5 m.
People who managed to hold on to their jobs through the bust have done quite well. With the housing bubble (is it a bubble?) causing a wealth effect and refinancing with low interest rates, a lot of people not only have a higher net worth but more disposable income as well.
BA is extremely well integrated into the world economy. For local workers this is a double edged sword. Booming economies in asia cause demand to soar, but jobs move quickly as well. The days of a good life in BA with a "IT" job may well be over, but skilled developers can still live well...ofcourse if paying 800k for a town house in Sunnyvale is living well...

[Jimbo]

IT is harder to oursource than programming.