As a starting point, consider how the rate at which new establishments of both kinds are being born has changed over time. Hathaway, Schweitzer, and Shane note: "As the figure shows, in 1978, Americans created 12.0 new firms per business establishment. By 2011, the latest year data are available, they generated new firms at roughly half that rate—6.2 new firms per existing business establishment. By contrast, in 1978, Americans created 1.7 new outlets per existing establishment, while in 2011 they created 2.6—an increase of more than half."
I wrote about about the decline of the top line--the line showing the entrepreneurs starting new firms--in "The Decline of U.S. Entrepreneurship" earlier this month.
Is this shift toward establishments started by existing firms just what one might call a "Walmart effect"--that is, big-box stores in the retail industry driving out smaller Mom-and-Pop operations? Apparently not. The share of new establishments that are new outlets of existing firms is rising across across all industries, not just retail
This change has implications for the sources of job creation in the U.S. economy. Hathaway, Schweitzer, and Shane write: "To give a sense of the magnitude of the changing sources of job creation, we can estimate the number of new jobs that new firms would have created had they continued to generate jobs at the rate they did back in 1978 and the number of new jobs that new outlets would have created had they continued to generate jobs at the rate they did back in 1978. At the 1978 rate of new firm job creation, new firms would have produced an additional 2.4 million jobs in 2011, or 90 percent more. At the 1978 rate of new outlet job creation, new outlets would have produced 828,000 fewer jobs in 2011, or 34 percent less."
As to underlying reasons for the change, these authors note that their underlying data doesn't allow them to pinpoint any particular cause. However, "[W]e can offer one hypothesis: Growth in information and communication technologies since the late 1970s have facilitated the coordination of multiple establishments, offering existing businesses an advantage over new firms when setting up new establishments to meet the need for new business locations."
This explanation seems plausible to me, but some other possibilities seem plausible, too. The new information and communications technology may also make it easier to establish brand names, so that firms in sectors like Finance, Insurance, and Real Estate or in Services are more likely to be part of a national firm, rather than opening up a stand-alone personal shop. I also suspect that the regulatory burden of starting a business has gotten harder over time, in terms of the rules, regulations, and permits that need to be followed for the physical property of the business, for dealing with employees, for meeting requirements about the qualities of the product that are provided, and for taxes and accounting. There's a case for each individual rule and regulation. But when you pile them all up, the burden can become discouragingly high for a potential entrepreneur.