Paul Romer and the S-Curve of Innovation 94.0

Theoretical economists and management theorists, like scientists viz engineers, often do not see eye-to-eye. The goal of the first is often new knowledge while those of the second is practical application. Their stakeholders, methodologies and measures of success are also different. Still, they often inhabit one S-curve though at different parts.

For this post, the S-curve I am trying to describe is of the business Innovation process, itself – the Solow (1958) black box or residual being opened and codified (from global, tacit) by venturesome economists and management theorists into a malleable process.

On my re-reading of Blue Ocean Strategy over the weekend , I am pleased to discover in Appendix C, The Market Dynamics of Value Innovation, the first use in a management book of the phrase, “nonrival and nonexcludable [sic] goods, such as knowledge and ideas,” words paraphrased out of the economist Paul Romer (1990), Endogenous Technological Change (See discussion in Posts #46 & #47).

The Blue Ocean authors also say in Appendix B that their reconstructionist view comes after Romer’s. I am inclined to believe that Blue Ocean is indeed an application of Romer’s theory along the S-curve of the business Innovation process.

Somewhere in SYNTHESiST, I did mention that the behavior of innovations over time is described by an S-curve (a Gompertz or logistic curve in math) and the typical time frame from invention to diffusion is 25 years. The typical innovation S-curve would have a flatter left tail than the sample on left, say for the first fifteen years, and a steeper up slope to fit 25 years. Points in a curve representing a major process, like business INNOVATION, could also be integrated from smaller s-curves representing incremental product and process improvements, say from 1982 JIT, TQC, re-engineering, business model, disruption, value migration, blue ocean, demand innovation, etc.

Blue Ocean Strategy was published in 2005. It is the first mention of Romer’s insights, I am aware of, in a very brief four-page Appendix of a major management book. It is a great milestone in the S-curve of the business Innovation process. In fact, its publication in 2005, 15 years after Romer (1990), did happen where the lower kink in the S-curve before it goes up the steep slope.

Below, I try a quick-and-dirty evaluation on whether the additional “factors of competition” introduced for some examples in the book do represent “new knowledge or ideas,” and non-rival and partially excludable thus leading to increasing returns. Also, I am analyzing this table to set-up initial hypothesis for determining general rules for screening competitive factors for best blue ocean effects. As such, these factors ought to allow network effects and increasing returns, the true source of the new blue ocean markets.

It would seem that each Blue Ocean strategic move does contain competitive factors (Yes/Yes) that tend towards network effects and increasing returns and, to me, are the true source of the blue ocean, unexplored markets.

I do believe endogenous technological change (as basic economic research) and blue ocean (as application research) reside in one S-curve for the business Innovation process. It could well be that Slywotzky’s Demand Innovation could be a fighter in the quiet battle for Dominant Design for this process.

It would be interesting to see if grounded theory can be developed from new hypothesis at setting rules as to what comprises new competition factors that work as we have tried in the table above. The (Yes/No) factors could well be enablers and necessary components of the successful strategy set.

When the theory is developed, then development of business models with network effects can go mainstream, i.e. Utterback’s dominant design is achieved (See Post #5). And the basis of competition thus changes …

Notes:

1. The 25 years typical time frame is the typical discovery-to-diffusion time frame of a major innovation. I have taken from either Utterback or Drucker, I am not sure now. Please email if you do know the original source.

Cost as Mirror Image of S-curves.(p. 85, Free, 2009 Chris Anderson0

2. From my library, the image on the left mirrors the steep part of the microprocessor S-curve – because shown as cost benefits of learning, it goes down, instead of learning which, goes up, showing the standard S. Note the clear s-pattern of the micro curves. This X-axis is just for 10 out of 25 years hence the flatter view. In this steep process improvement period, chip line widths were brought down from 1000 to 45 nanometers. One fab plant to produce chips at such narrow widths now costs in the US$billions. Click this link for our Post #93 on FREE by Chris Anderson.

3. The Blue Ocean authors use “nonexcludable” in the quoted phrase as in the image above; Romer actually referred to ideas and new knowledge as partially excludable. This is a critical distinction as Blue Ocean actually adds on new “competition factors” on existing products or services, i.e. already excludable goods by features or by brand equity. Thus, Romer’s original formulation partially excludable does apply. The exclusion portion is what locks in the values earned from the unexplored markets with the proponent firm.

4. The publishers need to update the Notes to Appendix C that cites Romer (1986) as the source of the “nonrival, nonexcludable” nature of technical change. These were introduced in Romer (1990) that is listed in the Bibliography. Romer (1986) was still founded on Solow (1958) where technical change is a residual in growth accounting.)

5. I have been guilty of similar lack of rigor as writer or teacher even as late as the immediately previous Post #93. I have been using and re-studying JIT-manufacturing but I always end up talking about the Breakeven instead of the EOQ diagram – as Schonberger’ book used – in teaching the theoretical proof of JIT. I do think, as technique of pedagogy, I can get by with a special case where EOQ and Breakeven can be equivalent. For example, a single-product company say in ketchup manufacturing moving to automated clean-in-place (CIP) from a manual cleaning process can be justified as initiating JIT with the EOQ or Breakeven showing equivalent curves.

Blue Ocean is a remarkable theory and I can forgive the authors and publishers for above (as long as they correct them in future editions).