The standard narrative about the American economy in the late twentieth century typically runs as follows: after four decades of quasi-statist capitalism (the 1940s through the 1970s) that brought the United States to the brink of economic ruin, the economy experienced an entrepreneurial explosion around 1980. The economy moved from the dull bureaucratic capitalism of the "Organization Man" to the exciting entrepreneurial capitalism of high-technology and, accordingly, more robust growth and productivity. Without questioning the importance of entrepreneurship to economic growth, I'm beginning to wonder about this.
For starters, let's take recent technological history. The information technology revolution that is associated with the 1980s and 1990s was born in the 1950s and 1960s, emerging out of government-funded research performed at universities and large company laboratories. We could probably mark the start of the IT revolution at the founding of Fairchild Semiconductor 1957 or, perhaps, the founding of Intel in 1968; Microsoft and Apple followed in the mid-1970s. Other companies typically associated with this entrepreneurial rebirth include FedEx and Southwest Airlines, which were also started in the late 1960s and early 1970s. Even within the confines of the conventional narrative outlined above, we can trace the roots of entrepreneurial capitalism to the final stages of bureaucratic capitalism, a pattern that aligns with other models of technological revolutions and transitions. In his wonderful book, The Age of Abundance, Brink Lindsey (who is now a Senior Fellow at the Kauffman Foundation), connects the counter-cultural ethos of the 1960s with the subsequent entrepreneurial revolution in business.
The typical narrative of entrepreneurial capitalism is also usually supported by reference to either changing rates of overall entrepreneurship or the high number of initial public offerings (IPOs) in the 1980s and 1990s, as well as the explosion of alternative forms of financing, from junk bonds to venture capital. And, indeed, according to at least one strand of Census Bureau data, the absolute number of new incorporations did increase dramatically from the 1960s to the 1970s and again to the 1980s and 1990s. New company listings, moreover, rose from an average of 156 per year from 1973 to 1979, to 549 per year from 1980 to 2001. From 1980 to 2001, in fact, newly listed firms comprised 10 percent of all listed firms in any given year. Venture capital, too, increased by several factors from the end of the 1970s and despite reverting from its astronomical heights in the late 1990s, VC remains at historically elevated levels.
Over the last several months, however, I happened to come across several pieces of evidence--some new, some older--that would appear to contradict the typical narrative. (I am not claiming any particular insight here, nor do I intend to mean that the U.S. economy was somehow "better" in the 1960 and 1970s than today. I heartily endorse Brink's arguments in "Nostalgianomics" on this point.)
Let's start with the most basic empirical information: the number and rate of new business formation. When the Census data cited above are expressed in per capita terms, the apparent doubling of firm formation disappears: the population doubled, bringing new business creation along with it without a corresponding increase in the rate. Indeed, we have published two papers on the puzzle of relatively steady levels of entrepreneurship in the United States, both across the past thirty years (the purported period of extreme entrepreneurship) and, potentially, across the entire twentieth century. In fact, firm formation fell from the 1980s to the 1990s and then flattened, neither falling nor rising. Per capita rates of new business formation today are actually quite similar--and certainly no higher--to those in the 1940s and 1950s, the ostensible high tide of bureaucratic capitalism.
What about IPOs? Using historical data from Gompers and Lerner and more recent data from Fama and French, here are the average annual IPOs for various time periods back to 1935 (the groupings are dependent on the authors' work, not my own manipulations):
The averages obviously mask variability, but it is generally true that after 1980 the United States saw sustained levels of historically high numbers of IPOs. But it is certainly not true that the era of bureaucratic capitalism was without entrepreneurial excitement. From 1959 to 1962, the average number of IPOs per year was 224. From 1968 to 1972 it was 388, a number much closer to the supposedly go-go 1990s than even the 1980s.
At the same time, the IPO explosion after 1980 brought its own issues. Fama and French found that the "decline in the cost of equity capital for new lists" meant that weaker firms could now go public. In particular, in the 1980s and 1990s the distribution of revenue growth across newly-listed firms skews to the right, meaning higher growth. Meanwhile, the distribution of profitability skewed left, meaning lower rates of profitability. Survival rates of newly listed firms also fell over this time period. As they stated: "Basically, the evidence says that weaker firms and firms whose expected payoffs are further in the future become viable candidates for public equity financing." This echoes other research finding, since the late 1970s, a dramatic rise in growth rate volatility among publicly-traded companies but a fall in the volatility of privately-held firms--nearly to the point of convergence, in fact. It is true as well that volatility among our largest companies, the Fortune 500, has risen and not quite one-fifth of today's Fortune 500 companies have been founded since 1980. But this is more an indication of turnover among existing companies, including mergers and acquisitions, rather than new firm entry. And even if more companies fall off the list, it doesn't mean they fail; it just as likely reflects the relative rise of other companies.
(Some of this also calls into question the role of high-growth firms. As the above comments indicate, rapid revenue growth does not necessarily mean profitability. This corroborates other work that examines the Inc. lists of fast-growing companies. I am a huge fan of the Inc. data and I think much more can be done with it, but we shouldn't imagine that high growth in revenues tells us anything about economic impact or long-term success. Presumably it should, but it apparently does not. I have written elsewhere that high-growth firms are neither as interesting nor important as they are typically thought to be. If the underlying dynamics of firm formation and competition are reasonably healthy, we should expect to see high-growth firms as epiphenomena of those, not necessarily drivers in themselves.)
In any case, what does any of this mean? Let's review. During the period of entrepreneurial capitalism, we have seen no increase (yet no significant decrease either) in levels of firm formation, but a larger number of IPOs and greater volatility among publicly-traded firms, indicated particularly by dispersed growth rates and profitability. This dimension in the public markets is, in part, an effect of greater availability of financing. Is what we call entrepreneurial capitalism really just a reflection of public-market volatility, itself a reflection of the rising role of finance in the American economy over the past thirty years?
I should point out, again, that none of this is meant to imply that the last thirty years were somehow worse than the middle part of the twentieth century. Nor is it meant to exclude the dramatic changes in technology over the past three decades. Indeed, high-tech companies likely account for much of the public-market volatility. At the broader level of aggregate firm formation, sectors such as food service, administrative services, health care, and professional services (in particular lawyers) tend to predominate, particularly among high-growth firms. My points here are not a value judgment on the types of companies that are formed. But it appears that there is still quite a bit about entrepreneurship--its trends, its composition, its impact--that we do not understand. And, should we be on the cusp of an entrepreneurship boom--whether because of technology, changing economic factors, demographic factors, whatever--it remains to be seen whether, in the context of existing structural factors, this will shift us to a different neutral state. And, again, we have not even touched the impact of new companies yet; a much more extensive treatment than I have given here will also layer on technological change. So much goes on beneath the surface of aggregate entrepreneurship levels, of IPO activity, of growth and profitability, that it is difficult to make blanket claims about capitalist typology.
I could, of course, be wrong.