Introduction: Does The "S" Curve Explain Anything About American Technological Innovation?

I always enjoy listening to the historian Doris Kearns Goodwin when she appears on NPR or MSNBC to discuss her historical interpretation of American presidents. I like listening for the underlying tension in her presentation about the appropriate timing for a historian to enter her judgment of history on a contemporary president.

The radio and television hosts are always trying to goad her and trick her into offering her historical judgments on the current occupant of the White House prematurely. This is like the delicacy of talking badly too soon about a recently departed, exactly how long does a person need to wait before offering their judgment of history?

Goodwin always obliges and delights her liberal listeners by opening up her historical kimono just enough to let her audience see that she thinks the "judgment of history" may not be kind to the current occupant of the White House.

I had been thinking about this topic in my deliberations on how the concept of the S curve of technology could be used to offer a judgment on the collapse of the American economy. My dilemma, like Goodwin’s, is that it may be too soon to offer a definitive economic assessment.

There are still many economic forces actively at work outsourcing the technological innovation capacity of the American economy. Even with the help of the S curve, it may be too soon to write the economic analysis of what went terribly wrong for America’s experience in the free trade global economy.

I wondered if following the S Curve backwards in time could be used to explain why outsourcing innovation caused the American economy to lose its ability for radical product innovation. Using current economic theory with supply and demand curves and marginal rates of efficiency somehow seems inadequate if the end state to be explained is economic extinction and not equilibrium.

In evolutionary theory, what currently exists today could easily have been something else if some earlier genetic crossover had occurred. For example, would the S curve be useful in predicting where the American economy would have been in 20 years from now if the multi national corporations had not begun outsourcing America’s technological innovation capacity around 1985?

Or, alternatively, can the S curve be used to look back in time to explain how the early decisions, around 1985, to outsource corporate innovation to India began the American economic decline? The series of speculative bubbles since then has tended to mask the underlying fundamental weakness of the economy, which is related to inadequate rates of domestic innovation and, consequently, inadequate market demand in America’s internal domestic industrial supply chains.

Perhaps, it would be enough simply to use the S curve to explain why the vacuous phrases like "transition to a knowledge economy," or remaining "globally competitive," are such barren platitudes about the disastrous effects of free trade when it is the nation’s capacity for innovation that is being freely traded away.

In the absence of domestic technological innovation, America’s initial factors of production, its cultural values of innovation and individual initiative, can not perform their economic function of creating new markets. New markets result from radical innovation, and cause the creation of new streams of wealth and new paths of upward occupational mobility for citizens at the lower ends of the income ladder.

Following the S curve backwards may help in identifying the time when the citizens of America began their transition from working in an innovation economy to subsisting in a third world welfare state.

Following the S curve forward provides clues to the much need economic revitalization of technological innovation in America, especially the metro regional economies that are being gutted by free trade.

What Would The S Curve Say?

In their research article, "Technological Evolution and Radical Innovation," Ashish Sood and Gerard J. Tellis (Journal of Marketing, Vol. 69 July 2005), begin by mentioning the conventional wisdom associated with the S curve of innovation.

"Currently, the literature suggests," they note, "that a new technology seems to evolve along an S-shaped path, which starts below that of an old technology, intersects it once, and ends above the old technology. This belief is based on scattered empirical evidence and some circular definitions."

They conducted historical analysis on the emergence 14 radical innovations taken from four markets in order to examine the shape and competitive dynamics of technological evolution.

They were interested in finding answers to the following questions:

• How do new technologies evolve?
• Do they follow the S shaped curve or some other pattern?
• Are technological changes predictable?
• Is the rate of technological change increasing?

Among the many fascinating conclusions they made was that the conventional interpretation of the S curve is wrong. "The results," they noted about their study, "contradict the prediction of a single S-curve. Instead, technological evolution seems to follow a step function, with sharp improvements in performance following long periods of no improvement. Moreover, paths of rival technologies may cross more than once or not at all."

Their description of long periods of stasis, followed by outbursts of innovation, sounds similar to the new Darwinian interpretation of evolution that biologists call "punctuated equilibrium." The implications of S curve punctuated innovation for America is that if long periods of stasis are not interrupted by a burst of innovation there will not ever be future bursts of innovation because the genetic technological diversity in America never occurred.

It occurred somewhere else, as a result of the MNCs outsourcing innovation, primarily to India and China.

For Sood and Tellis, "In nine technologies, we did not find a single S curve. Rather, we found long periods of static performance interspersed with abrupt jumps in performance. The plots suggest a series of step functions, each of which could approximate an S curve."

Another challenge to conventional wisdom they found was that radical innovations occur in really big corporations, like the American multi national corporations that have been outsourcing innovation since 1985.

"In contrast to the dominant view in the literature (H8)," they stated, "we find only 1 platform innovation introduced by small entrants. All the remaining 13 platform innovations came from large firms (7 incumbents and 6 new entrants). Although our results run counter to the dominant view in the literature, they are consistent with two recent findings in the literature (Chandy and Tellis 2000; Sorescu, Chandy, and Prabhu 2003)."

Even with their deep finanical pockets for radical innovation, it takes the large corporations about 15 years to run through a radical innovation cycle. In terms of what they call gestation time, they concluded, "We also examined the gestation time of each technology, which is defined as the time it takes for a firm to convert a patent to a commercial product. The average gestation time for technologies is 14.5 years for display monitors, 14.3 years for desktop printers, 9.7 years for desktop memory, and 22.7 years for data transfer technologies. The overall average for all categories is 15.1 years."

To provide some historical perspective, the innovations that occurred around 1985 have already run their course, ending up as obsolete products in the junkyard of the global market.