Silicon Valley Nation: Semiconductor jobs problem

SAN FRANCISCO -- New employment numbers out from the Semiconductor Industry Association (SIA) this week are meant to paint an encouraging picture of the importance of semiconductor employment to the U.S. economy.

In November, the SIA reported that the chip sector is adding jobs faster than the U.S. economy. The semiconductor industry’s manufacturing workforce grew by 3.7 percent over the past year. In comparison, jobs throughout the broader U.S. economy increased by 1.2 percent over the same time period.

This week, the SIA trickled out more data showing the impact of semiconductor innovation on state-by-state employment.  California leads all states with 47,100 semiconductor jobs, followed by Texas (28,800), Oregon (23,400), Arizona (18,800) and Massachusetts (10,100). New York ranks sixth among all states with 7,600 semiconductor jobs. Rounding out the top 10 are Idaho (7,400), Florida (7,100), Vermont (5,100) and New Mexico (4,500).

All employment figures reflect recently released 2011 data from the U.S. Bureau of Labor Statistics (BLS).



SIA's doing a laudable job in Washington of using data like this to pound the industry's drum a little louder, but there's a problem with these numbers that the SIA can't really control. The employment figures tell us what we already know and have known for almost 30 years: That our industry is concentrated in a few key states. You might as well call them the State of Intel, or the Commonwealth of IBM, right?

The question is why hasn't the industry since the 1980s made a wider impact on local economies?

Empty spaces?

There have been some outliers: Cypress Semiconductor has a plant in Minnesota. OK. But in a nation with disparate cost bases, a world-class college system and an excellent transportation system, why haven't we seen more manufacturing or design huddle in Iowa, North Dakota, Georgia or Louisiana?

• Some of it is habit (expand where there's an existing semiconductor infrastructure).
• Some of it is moving offshore to emerging markets and global partners.
• Some of the onus is on local states for thinking, for years, that the chip industry was about potatoes.

What I'd love to see is an analysis (and it would take a lot of effort) of the BLS numbers to get a sense for how the rise of the Internet has spawned many more (small and dispersed) design teams in non-traditional states. Anecdotally, we saw a lot of this on our year-long Drive for Innovation--system design is happening everywhere. Some of those small operations will grow into the next Broadcoms or Marvells of the world. For now, they remain under the radar.

What these numbers also tell us is that the big impact of the semiconductor industry on U.S. employment is still isolated to headquarters locations and clusters of manufacturing regions.

If we're worried about jobs moving offshore, there's still plenty of cost-effective and infrastructure-ready regions in the U.S. we need to consider.

Related stories:

--Silicon Valley Nation: Lobbyist optimistic on fiscal cliff
--Silicon Valley Nation: 30-year battery life? Really?