The following is a guest commentary for CNBC.com.
There is a certain traditional narrative of the startup. Smart kid comes to Silicon Valley, thinks of a billion-dollar idea, gets worldwide customers, and finally fulfills wish of diving into a pool of money Scrooge McDuck style.
Of course, the narrative is much more complex. Startups fail at a high rate, pivoting occurs often, and diving into a pile of coins will quite likely cause grave bodily harm. But the part that many do not question is the need to go to Silicon Valley or some other large startup hub, such as Boston or Seattle.
But it turns out that the chronicle of where people choose to start companies as compared to where they come from is far more complex and much more interesting than you might imagine.
Recently at the Kauffman Foundation, we used data from the Inc. 500 to examine where founders went to school versus where they founded companies and examined how people move about the country, at least when it comes to startups.
And we got a picture that is far from the simple Silicon Valley narrative.
We did find that the Bay area and Boston educate many founders, and even retain many of them. (More: 11 Notable Entrepreneurs Teaching the Next Generation)
But we also found that many people start companies in places we might not expect, such as the South, and the cities that had the highest net flow (the biggest gains and retentions of founders, minus losses) is topped by Washington, D.C., Atlanta and Dallas.
When it comes to the highest net loss, college towns feature prominently, as they educate many founders who then start companies elsewhere. While this includes such places as Ann Arbor and Urbana-Champaign, San Jose—the home of Silicon Valley—features prominently on this list as a city with a high net loss.
Of course, some traditional cities did top the lists of where founders are attracted, with the top five being D.C., Los Angeles, New York, San Francisco and Atlanta. But overall, the picture is more nuanced than everyone simply going to the Bay area.
But even more than that, we found that many people tend to stay within the region where they received their degree. Those in the Northeast tend to stay in the Northeast, and those in the South tend to stay in the South. (More: What Is the Worth of an Ivy League Education?)
Furthermore, we also used a network analysis to see which metropolitan areas are connected to each other based on migration. And we found that many of the communities in this picture are those defined roughly geographically — people move between places that are relatively close to each other.
So, while many people proclaim the "death of distance" with the advent of the Internet and the ease of telecommuting, location still clearly matters. And where you start (or at least where you go to school) can exert a profound impact on where you end up starting a company, whether or not it yields a pile of gold you can swim in.
Samuel Arbesman is a senior scholar at the Ewing Marion Kauffman Foundation and the author of the new book, The Half-Life of Facts.