Small business activity does not measure entrepreneurship

One of my papers has been published in Proceedings of the National Academy of Sciences (PNAS). Magnus Henrekson and Tino Sanandaji ”Small business activity does not measure entrepreneurship“:

“Schumpeterian entrepreneurship refers to growing and innovative firms. However, in empirical research the rate of entrepreneurship is commonly estimated using the self-employment rate or other measures of small business activity. We argue that this empirical strategy gives rise to misleading inferences regarding Schumpeterian entrepreneurship. To unambiguously identify this type of entrepreneur we focus on self-made billionaires on Forbes Magazine’s list who became wealthy by founding new firms. We identify 996 such billionaire entrepreneurs in over 50 countries. The rate of billionaire entrepreneurs correlates negatively with self-employment, small business ownership, and startup rates. Countries with higher income, higher trust, lower taxes, more venture capital investment, and lower regulatory burdens have higher entrepreneurship rates but less self-employment.”

The Economist online entrepreneurship section “Schumpeter” writes about our study What exactly is an entrepreneur?:

“Most small companies are Mom-and-Pop stores that will always stay in the family. Three-quarters of people who start companies say that they want to keep their companies small enough to manage themselves.

In a new paper Magnus Henrekson and Tino Sanandaji argue that the number of self-made billionaires a country produces provides a much better measure of its entrepreneurial vigour than the number of small businesses. The authors studied Forbes’s annual list of billionaires over the past 20 years and produced a list of 996 self-made billionaires (ie, people who had made their own money by founding innovative companies as opposed to people who inherited money or who had extracted it from the state). They demonstrated that “entrepreneur density” correlates with many things that we intuitively associate with economic dynamism, such as the number of patents per head or the flow of venture capital.

They also demonstrated it correlating negatively with rates of small-business owners, self-employment and startups—in other words that many traditional measures are about as misleading as you can get. Countries with a lot of small companies are often stagnant. People start their own businesses because there are no other opportunities. Those businesses stay small because they are doing exactly what other small businesses do. The same is true of industries. In America industries that produce more entrepreneur billionaires tend to have a lower share of employees working in firms with less than 20 employees.

This makes sense: successful entrepreneurs inevitably destroy their smaller rivals as they take their companies to scale. Walmart became the world’s largest retailer by replacing thousands of Mom-and-Pop shops. Amazon became a bookselling giant by driving thousands of booksellers out of business. By sponsoring new ways of doing things entrepreneurs create new organisations that employ thousands of people including people who might otherwise have been self-employed. In other words, they simultaneously boost the economy’s overall productivity and reduce its level of self-employment.

Who are the Schumpeterian entrepreneurs who dominate the modern economy? And how do you create more of them? Messrs Henrekson and Sanandaji argue that the majority of the world’s wealthy entrepreneurs acquired their riches by starting a business: 65% in America, 42% in Europe and 52% overall….

The authors are less informative about the second question. They warn that high taxes can encourage replicative entrepreneurship rather than innovative entrepreneurship. The self-employed face lower tax rates than the employed (and can evade taxes more easily). They also face a lower chance of being audited. This encourages companies to stay small and encourages workers to sell their labour to small companies rather than big companies. The same is also true of heavy regulation. They warn that conceptual confusion over the nature of entrepreneurship can also create policy confusion: attempts to boost the number of small businesses can reduce the likelihood that one of those small businesses will outcompete all the others.

Schumpeterian entrepreneurship is all about innovation and ambition to turn small businesses into big ones. Small business entrepreneurship is all about flexible employment and poor opportunities. But the authors have little to say about how to create the network of institutions that they think helps to create entrepreneurship: high-powered universities and dense clusters of activity of the sort that flourish in Boston and Silicon Valley.

Still, if Henrekson and Sanandaji do not provide us with the key to the secret kingdom, they at least make sure that we are trying to get through the right door.”

The Economist online blog on American politics also writes about our study Apple carts or Apple?:

“Entrepreneurs come in two categories. The first are self-employed types who run small businesses that replicate other businesses (eg, nail salons, dry cleaners, etc). The second are disruptive innovators who come up with new ideas. The first type of entrepreneur creates some jobs, but the innovative entrepreneur contributes disproportionately to growth—Apple will expand more than an apple stand. And here’s where it gets interesting: it turns out that high-growth entrepreneurs are more common in places with lower levels of self-employment, according to a new paper by Mangus Henrekson and Tino Sanandaji (which will be discussed in greater detail on our Schumpeter blog).

This is actually less mysterious than it seems. Self-employment is high in places where it is hard to expand a business. This is why people are more likely to work for themselves in Greece, Italy, Turkey and Portugal—places where people also report a low level of social trust. Innovation requires a willingness to bet on something risky, and a faith that a good idea will be rewarded. Few are likely to take this leap in places where people and banks are viewed with suspicion. This is why a high level of self-employment correlates with a low level of disruptive entrepreneurial activity.
In a ranking of OECD countries, America has the second-lowest number of self-employed people, and the second-highest number of entrepreneurs per million inhabitants, after Israel. Yet America is not an especially high-trust country: only a little over half (52%) of Americans view people as “generally trustworthy”. But as we have argued before, America does not behave like a low-trust country, either. It is trusting enough, the authors think, that its other advantages, like a sophisticated financial system and superb universities, foster lots of innovation (particularly in Boston, New York and Silicon Valley).”