Why Silicon Valley's Tech Culture Hasn't Moved to Texas
(Bloomberg View) -- Brooklyn-born Bob Metcalfe got his tech-industry start where the U.S. tech industry got started, in the Boston area. After studying at the Massachusetts Institute of Technology and Harvard University and working on MIT's Project MAC, a pioneer in operating systems and artificial-intelligence research, he headed west in 1972 to work at the now-legendary Xerox Palo Alto Research Center. There he led the invention of the crucial computer networking technology known as Ethernet, after which he co-founded 3Com, the networking-equipment maker that was for a time one of Silicon Valley's leading lights (it was acquired by Hewlett Packard in 2010). Now, he spends most of his time in Austin, where he is professor of innovation and Murchison Fellow of Free Enterprise at the University of Texas.
So when Metcalfe saw a certain New York Times headline earlier this month about Silicon Valley being "over," he had a perhaps understandable reaction:
I wondered about that, though. Yes, housing is much more affordable, taxes are somewhat lower and the roads are marginally less clogged in Texas than in California. Also, the Texas economy has undeniably been booming, creating 3 million new jobs since 2000, compared with 2.7 million in much more populous California.
But I remembered looking through what is now called the PwC/CB Insights MoneyTree Report a couple of years ago and being surprised at how little venture capital Texas attracts relative to Silicon Valley. When I looked at the MoneyTree report again and added up the numbers for the past five years, I was even more surprised to learn that the state hadn't even attracted as much VC funding as the District of Columbia and its environs:
Texas' share of U.S. VC investment used to be bigger. It averaged almost 6 percent from 1995 through 2012, compared with just 2.2 percent over the past year. Meanwhile, California has gained share since the 1990s and 2000s, and metro Boston (which accounts for the overwhelming majority of New England VC investment) and metro New York have gained share lately.
Texas-based companies used to be much bigger players in the tech world, too. In the late 1950s, the integrated circuit, which made the modern tech industry possible, was invented at Texas Instruments in Dallas before being subsequently reinvented and improved at Fairchild Semiconductor in Silicon Valley. Later TI popularized the pocket calculator and led the emergence of the Dallas area as a major tech hardware manufacturing hub. In the 1960s and 1970s, Ross Perot's Electronic Data Systems, based in the Dallas suburb of Plano, helped bring computerization to corporate America. In the 1980s and 1990s, Compaq (Houston) and Dell (headquartered originally in a University of Texas dorm room, later in the Austin suburb of Round Rock) were leaders of the booming personal-computer industry. Dallas-based Sevin Rosen Funds, founded in 1980, invested in Compaq, Citrix Systems (which was founded in Texas in 1989, but moved to Florida a few months later), Electronic Arts and Lotus Software on the way to becoming one of the country's top VC firms. And in 1999, Broadcast.com, also in Dallas, was one of the first companies to cash in big on the internet boom, selling out to Yahoo! for $5.7 billion and ensuring that co-founder Mark Cuban could afford to keep himself in the media spotlight for the rest of all of our lives and probably beyond.
TI and Dell are also still around, but TI is much diminished and Dell somewhat so. Compaq and EDS were both eventually swallowed up by Silicon Valley-based HP (EDS has since been spat back out again as Virginia-based DXC Technology). Sevin Rosen hasn't raised a fund since 2004.
Meanwhile, on CB Insights' current list of "unicorn" startups with valuations of $1 billion or more, only two of the 117 U.S. companies -- mobile commerce software maker Mozido and Kendra Scott Jewelry, both based in Austin -- are from Texas. Seventy-one are from California, 16 are from New York, and seven are from Massachusetts. Of CB Insights' list of 118 billion-dollar-plus exits since 2009, Texas can also claim just two -- Austin-based vacation rental site HomeAway, which went public in 2011 and was acquired by Expedia in 2016, and Pasadena-based biofuels maker KiOR, which went public in 2011 and bankrupt in 2014 -- to 68 for California, seven each for Massachusetts and New York, and six for Washington. Finally, if you prefer semi-obscure pop-culture references to hard data: The first two seasons of the underappreciated AMC historical tech-startup drama "Halt and Catch Fire" were set in Dallas, then the main characters moved to Silicon Valley for the last two. Need I say more?
Actually, I probably do need to say more. Texas is a state of big spaces and big companies, with 50 Fortune 500 headquarters, just behind New York's 54 and California's 53, and lots of other major corporate operations. Maybe the focus on venture capital and startups is misplaced, and tech employment is a better gauge.
The tech sector that employs the most people in the U.S. is what's called computer systems design and related services -- the territory of such firms as Deloitte, Accenture, the aforementioned DXC Technology, and the services arm of IBM. It's big and getting bigger in Texas, with the state's share of U.S. employment in the sector growing from 4.6 percent in 1990 to 8.3 percent now. California's 15 percent share is about the same as what it was in 1990 and 1999, although it went higher in between. New York has lost the most employment share.
Another big tech sector in which Texas has increased its share of U.S. employment, from 6.5 percent in 1990 to 8.5 percent now (it got up to 9.5 percent in 2012 and 2013), is in computer and electronic product manufacturing, where jobs range from low-paid assembly labor to super-high-end semiconductor design work (overall, weekly earnings in the sector averaged $1,466 in February). But that isn't because Texas has been adding jobs, at least not since 2001. It just hasn't been shedding them quite as fast as the rest of the country.
So Texas has been gaining ground in (1) tech manufacturing, which is shrinking, and (2) computer systems design and related services, which is a growing sector that offers good jobs -- national average weekly earnings were $1,811 in January -- but does not really occupy the commanding heights of the tech industry.
What does occupy those heights? Well, there's software publishing, which includes large parts of Microsoft, Oracle, Adobe Systems and the nation's video-game makers. I say "parts" because the BLS measures payroll employment by establishment, meaning that employment at each physical location of a company is supposed to be classified according to what most of the people there do. So a Microsoft Azure data center would likely fall under the separate category of data processing, hosting and related services, while an Oracle Consulting office might be under computer systems design and related services. Jobs in software publishing pay an average of $2,199 a week nationally, $2,547 in Texas, $4,320 in California and $6,706 (!!!) in Microsoft's home state of Washington. Those state numbers are for the third quarter of 2017, and they're from the BLS's Quarterly Census of Employment and Wages, which isn't as timely as the monthly payroll employment report but contains state and local data on smaller sectors such as software that the payroll reports don't.
So Texas has been losing ground in software publishing employment to California, Washington and even Massachusetts. Then there's "internet publishing and web search portals," which would seem to cover most of what Facebook and Google do. It also covers the not particularly high-tech or high-status work of journalists at online-only publications, but given that average weekly earnings in the sector in California were $5,609 in the third quarter of 2017, it seems like it must be mostly places like Facebook and Google there. Here's employment by state in the category, which has only existed since 2007. Texas didn't even make the top five, so I made the chart go to six.
And here's employment in data processing, hosting and related services, which isn't all that huge or glamorous (national average weekly earnings: $1,521; it is $3,375 in California) but is a category for which the BLS has monthly data going back to 1990.
There is another way of measuring tech employment: by using annual estimates from the BLS's Occupational Employment Statistics program, which tracks what people do rather than where they work. Because of the way the survey is conducted, and because occupational categories change over time, the BLS discourages making time series out of the data. But here are a couple of snapshots: In 1998, California and Texas were Nos. 1 and 2 in the number of people working in "computer, mathematical, operations research and related occupations," with California's share of the national total at 11.1 percent and Texas' at 8.2 percent. In 2016, the two states were still 1 and 2 in the similar category of "computer and mathematical occupations," with California's share 13.9 percent and Texas' 8.4 percent.
It doesn't look from all this like Silicon Valley is moving to Texas! When I shared a synopsis of these findings with Metcalfe, though, he responded that, well, lots of actual Californians are moving to Texas. He's right about that. The California Legislative Analyst's Office put out a useful summary of domestic out-migration data last month. Texas is by far the leading destination for fleeing Californians, with a net inflow of close to 300,000 from 2007 to 2016. And while most states have experienced net inflows of lower-income and middle-class people from California but net outflows to California of those making more than $200,000 a year, California's migration deficit with Texas spans the income spectrum (although it is most pronounced among those making $55,000 or less). California is also losing residents in the prime career age range of 26 to 45 to Texas, even as it gains people in that age range from Illinois, New York and even Florida.
Still, California experienced even bigger out-migration waves in the early 1990s and early 2000s, and Silicon Valley seemingly just got stronger. Also, the outflows over the past decade, according to another useful California Legislative Analyst's Office report, have been overwhelmingly from Southern California. Santa Clara County, the heart of Silicon Valley, actually saw net inflows from other states and counties from 2007 to 2016, although it's possible the flows have reversed more recently.
When the epicenter of the U.S. tech industry shifted from Route 128 in the Boston suburbs to Silicon Valley in the 1970s and 1980s, it wasn't because the real estate was cheaper, the taxes lower or the roads less crowded in California than in Massachusetts. In fact, I'm pretty sure the real estate was already more expensive and the taxes similar, although the driving in those days was probably less stressful in California, and the weather was and is definitely better. But the main reason for the shift was that, as AnnaLee Saxenian explained to me a few years ago:
The Boston area was organized around these big, vertically integrated minicomputer companies — DEC, Data General. They were classic postwar American companies, with vertical hierarchies and career ladders. Planning and research happened at the top of the organization and then funneled down. Whereas in Silicon Valley you had, really by chance not design, a series of flat companies, with project-based teams that moved around. People moved between companies much more fluidly. At a time that technology and know-how were sort of trapped within the vertically integrated companies of Route 128, they were being continually recombined in Silicon Valley. That gave them a real edge in innovation.
Saxenian is dean of the School of Information at the University of California at Berkeley and author of the classic comparative study "Regional Advantage: Culture and Competition in Silicon Valley and Route 128." To apply her analytical lens to Texas: The state's early tech companies were big, career-ladder-oriented kinds of places, but they weren't as vertically integrated as DEC and Data General, and they did spawn startups. Compaq, for example, was founded by a trio of Texas Instruments engineers.
Another factor that has been cited for Silicon Valley's success is that noncompete clauses are almost impossible to enforce in California, meaning that there's not much companies can do to stop employees from taking their expertise to startups. Interestingly, according to Dallas attorneys W. Gary Fowler and Sarah Dobson Mitchell, noncompete law in Texas has shifted from quite permissive in the late 1980s and 1990s to extremely tough on departing employees since 2006.
This fits with a general sense I have that, as Texas has attracted more and more big corporations to locate headquarters and major operations there, it has developed a more big-corporation-oriented economy -- which can raise hurdles for startups. Texas universities, venture capitalist Kevin Stevens told me, tend to steer their best students toward jobs at American Airlines, Dell or Exxon Mobil, not at new companies.
I had contacted Stevens, a partner at Intelis Capital in Dallas, because he had written one of the only pieces I could find that described the Texas tech/startup scene as anything less than booming, a blog post last summer that was subsequently picked up by VentureBeat and headlined "What’s keeping Texas from rivaling Silicon Valley."
What's keeping Texas from rivaling Silicon Valley, Stevens wrote, is a lack of funding, mentors and collaborative institutions. "I think there was for 10 to 15 years a gap of good VC funding," he said when I got him on the phone last week. Venture capitalists in Silicon Valley and New York generally insist that startups in which they invest move to Silicon Valley or New York, so it takes Texas-based venture capital to build a startup culture there. And while Stevens said the Texas startup-investing scene is now on an uptrend, it has a long away to go.
He's also doubtful that the refugees from California are going to jump-start things. "There is an influx of technical talent, but I'm not so sure that's the technical talent that wants to start companies," he said. "If you make $150,000 as an engineer in the Bay Area, you can barely afford to live there. A $150,000 job here gets you a great three-bedroom or four-bedroom home, and there's no reason to start a company."
So what could turn Texas tech around? "We actually think Texas is well positioned to disrupt some of the industries in most desperate need of disruption, such as energy and health care," Stevens said. "The Bay Area is great at disrupting industries where naivete helps," he added, while in energy and health care there are regulations and technical challenges that make industry expertise more important -- and Texas is the headquarters of the energy industry and boasts several leading centers for medical research. In these fields, in fact, Texas is already a high-tech power. As Metcalfe put it in an email:
Texas is where two of the most impactful new technologies are saving the USA’s butt right now. Horizontal drilling and hydraulic fracturing of oil and gas.
Yes, but ... that still doesn't mean Silicon Valley is moving to Texas.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Justin Fox is a Bloomberg View columnist. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”
Texas has much higher property taxes than California, but no income tax, while California has the highest income tax rates in the country Overall, according to the Tax Foundation California had a state and local tax burden of percent of GDP as of 2012, while in Texas the burden was percent. Meanwhile, the median sales price of an existing single-family home in the fourth quarter of according to the National Association of Realtors was million in metro San Jose and in metro San Francisco, while it was in metro Austin, in Dallas-Fort Worth and in metro Houston.
The region names in the chart are the ones that PricewaterhouseCoopers has been using for years in its VC reports. It did split the region it previously called Silicon Valley last year into San Francisco (North Bay Area) and Silicon Valley (South Bay Area).
TI had employees in 1990 it has 29,700 now
Jobs site Indeed, which had and has a big presence in Austin but was headquartered in Stamford, Connecticut, is also on the list for its billion 2012 sale to Tokyo-based Recruit Holdings.
The "supposed to be" is important here. The BLS urges employers to break jobs down by establishment for the monthly Current Employment Statistics and Quarterly Census of Employment and Wages, but employers aren't required to do it the BLS isn't allowed to disclose anything about which employers report what, which is why there's an element of conjecture to all of this.
For more columns from Bloomberg View, visit http://www.bloomberg.com/view.
©2018 Bloomberg L.P.