STARTUP

Stanford And Harvard Profs Explain Why These Two Regions Host Most Startups

If the world is flat why do most of the fastest growing startups locate in just a few regions? In a nutshell, agglomeration has its advantages.

Share on
BY Peter Cohan - 20 Jul 2017

PHOTO CREDIT: Getty Images

As someone who enjoys making money on my investments, I find it useful to consider where money-making investments come from.

My experience has been that luck plays a huge role. But that luck seems to be concentrated in certain parts of the world. More specifically, my most successful investments have been in startups from Boston and San Francisco that turned into shares of a publicly-traded company.

I am guilty of repeating the conventional wisdom that startups are concentrated in a small number of cities around the world. But in a July 7 interview, Harvard Business School Bruce V. Rauner Professor of Business Administration, Jan Rivkin, pointed out that this statement is not precisely true -- it depends on how you define startup.

If a startup is a small, private company -- such as a restaurant -- that is intended to help a founder and its employees to make a living, such startups are widely distributed geographically.

However, if a startup is thought of as a small private company whose goal is to get big fast -- whose investors hope to get richer when the company is acquired or goes public -- then startups are concentrated.

The Kauffman Foundation refers to such fast-growing startups as Gazelles -- emphasizing that such startups are distinguished by their ability to grow quickly while continuing to develop new products, win customers, and deliver a high level of service. Kauffman argued that Gazelles account for 50% of new jobs, they expand into new geographies, and they create growth in related industries.

Rivkin suggested that there were good reasons why Gazelles locate in relatively small number of specific cities such as Boston, Silicon Valley, Los Angeles. As he said, "Venture capital is spiky. The top 50 metro areas receive 97% of the venture capital. 83% of the venture capital investment goes to places like San Francisco, Boston, and Southern California."

Such concentration happens because it works. "It goes back to Alfred Marshall in the 1890s. Agglomeration happens for a reason -- there are positive feedback loops between skilled labor and specialized inputs such as venture capital. Computer scientists want to be in the Bay Area; biotechnologies flock to Boston and Cambridge; media people go to New York. As Marshall wrote, 'Mysteries are as if they were in the air,'" explained Rivkin.

Stanford Business School Fred H. Merrill Professor of Economics, Paul Oyer, argues that "it is almost impossible for cities that attempt" to make themselves into the next Silicon Valley. According to Oyer, "[Startup cities] get started thanks to research and education. Towns with noted universities like Silicon Valley, Research Triangle Park, and Austin get people who graduate, live there, and get together. The network is valuable."

Oyer points out that universities are a necessary but not sufficient condition. "If universities alone were all that was needed, Missoula, Mont. - where University of Montana is located - would be a startup hub. You also need companies. For example, Stanford had Hewlett Packard and Schlumberger."

Oyer sees many factors that could, but don't, destroy Silicon Valley as a startup hub. "It is hard for other cities to break in because of the power of the network that gets created which grows as each new person comes here. And venture capitalists want to be where the talent is. The strength of the network overwhelms all the factors that should kill Silicon Valley such as the high cost of living and high tax rates," he explained.

Ultimately, there is a very compelling reason why other cities want to host Gazelles -- they create tremendous local wealth -- boosting tax revenue and housing values and creating jobs. And the limiting resources for creating Gazelles are big increments of talented people and capital.

As Harvard Business School Professor of Management Practice Shikhar Ghosh said, "A company like Facebook which is trying to get big fast can get the resources it needs in Boston when it's small, but if it wants to get really big, it needs to move to a place like Silicon Valley where it can more easily raise the capital it needs in big increments. A company that ultimately want to employ 20,000 to 30,000 people will move to an area that has the capital to help it raise Series B, C and D funding."

Another key resource that Gazelles need to hire in big chunks is talented people. George Foster, Konosuke Matsushita Professor of Management at the Stanford Graduate School of Business, said "In Silicon Valley an entrepreneur hire 500 software engineers quickly. And China is becoming another such place -- it produces 200,000 engineers a year and 2,000 of them are truly great."

Harvard Business School Associate Professor, William Kerr, argues that culture and the desire of young talent to live in cities is also driving the concentration of Gazelles into a few locations. As Kerr explained, "Culture favors some places over others. People want to be in places where everyone idolizes entrepreneurs. They want to be where you can try and fail and not be ostracized. In the Boston area, young talent wants to live in cities which is making it hard for me to find a buyer for my three acre property in [tony suburb 25 miles from Boston] Lincoln."

I was happy to learn that not everyone thinks that Silicon Valley "embraces failure." George Foster, Konosuke Matsushita Professor of Management at the Stanford Graduate School of Business, said, "It is dead wrong to say that Silicon Valley 'embraces failure.' It is correct to say that Silicon Valley 'tolerates smart failure.' Failure is painful -- it costs investors money and people lose their jobs. If you failed because you were too early or more adventurous and you learn from it, then such failure [is acceptable]. If you fail because you were a doofus, it is not embraced."

The trend of Gazelle concentration does not seem to be slowing down -- but a look at the histories of Detroit or Pittsburgh tells us that Gazelles might gather in new locations over time.