Why ‘creative disruption’ is overblown

Brian Wallheimer | Jun 15, 2017

Sections Economics

Just a decade ago, Nokia phones accounted for half the high-end market, and the company was the envy of technology circles. Then Apple’s iPhone—and the Google Android phones by Samsung and others—toppled Nokia’s offerings. There are many similar stories: Priceline and Expedia undid travel agents, Uber has replaced many taxis, and Airbnb has challenged hotels worldwide.

The concept of disruptive innovation has become so ubiquitous, many economists say that creation and economic growth stem mostly from new companies upending entrenched ones. But the contribution of these disruptive companies to overall economic growth is relatively small, according to the International Monetary Fund’s Daniel Garcia-Macia, Chicago Booth’s Chang-Tai Hsieh, and Stanford’s Peter J. Klenow.

The researchers examined job-growth patterns using US Longitudinal Business Database records from 1976, when the database was formed, to 2013. The data encompass payroll records of all nonfarm firms in the private sector, giving the researchers a snapshot of the number of US jobs created and destroyed each year.

They compared 1976–86 and 2003–13 to see how much of the economy’s employment growth and GDP growth could be attributed to incumbents, and how much to new entrants, in these 10-year periods. The researchers argue that the role of entrants should be tied to the employment share of young companies. Creative destruction should show up as big declines among some incumbents (some of which even exit the market)—and major growth among the companies doing the destroying. That is, creative destruction should show up in the “tails” of the distribution of employment changes across companies.

In contrast, when incumbents improve their own products, they replace their own offerings, entailing much smaller changes in employment. Think of a retailer upgrading an outlet as opposed to opening a new store and driving out a competitor. Thus the effect of such incremental innovations within companies should populate the “middle” of the job-growth distribution.

Hsieh and Klenow conclude that creative destruction explains the extreme employment declines and company exits seen in the data. But most employment changes, even over five-year periods, are much more modest. Thus the researchers conclude that most innovation takes the form of incumbents improving their own products.

From 1976 to 1986, creative destruction accounted for 19 percent of the US’ GDP growth. The rest came from companies introducing new products and services to replace their own, or incrementally improving on their earlier offerings. From 2003 to 2013, creative destruction accounted for an even smaller portion of growth: only 13 percent of GDP growth came from new entrants displacing incumbents.

Forty years ago, economic growth wasn’t dependent on massive upheavals in the marketplace, the researchers write, and it is even less dependent on these shifts today. Over this 40-year span, only 10–12 percent of new job growth can be attributed to creative destruction.

“If you want to grow, the main focus is not so much about the big things; it’s about the small and little things you can do,” Hsieh says. “Home runs are rare. Companies grow by hitting single after single and steadily building.”