David Warsh introduced me to Paul Romer in his 2006 book, Knowledge and the Wealth of Nations. In the otherwise dismal field of economics, he succeeded in turning the good professor’s dreary work of writing a research paper into an epic voyage of discovery.
The plot spans a couple of centuries; the characters include the brilliant and the petty locked in intellectual combat; and the scene is in austere Italian classrooms or the hallowed halls of academia.
I think this work is worth a Nobel Prize for Professor Romer. Its genius (and I believe its future applications) comes because he had to create the tools to discover the insight as did Newton with calculus and physics.
The work was published as "Endogenous Technological Change" in the University of Chicago’s Journal of Political Economy in October 1990. Mr Warsh noted the directly identifiable value of Professor Romer’s work.
For the first time in the context of growth theory, [Romer, 1990] embedded [the concept of intellectual property] in an aggregate-level model of the economy, describing knowledge as both an input and an output of production in a way that permitted economists to take account of its significance.
Like most basic research, it stands on the shoulders of giants.
Its most direct antecedent is 1997 Nobel prize winner Robert Solow’s findings in 1957 that only 20% of American growth then can be explained in terms of the basic factors of production, labor and capital.
In 1957, the then state-of-the-art in mathematical economics, including the assumption of perfect competition, denoted the balance of 80% as a residual; the catch-all basket of growth coming from all kinds of ‘technical progress.’ It was also an externality to the economy.
From 1957, others have tried to account for technical progress to include those from learning by doing (Arrow, 1962), using (Rosenberg, 1982) and interacting (Lundvall, 1988). These later findings came mainly from inductive case study and historical approaches.
It took for Romer to adapt new mathematical tools to allow for a better accounting of Solow’s residual. From an externality, his 1990 model, in effect, brought technology inside as an engine of the aggregate economy.
Does the USA Have Small Government? In so doing, the model also eliminated the need for perfect competition and a fixed rate of technical progress as simplifying assumptions. The implications in policy and in politics are not yet reflected in current American debates on health care reform and on size of government involvement in the economy.
I think the model can better prove — by deriving the eventual productivity benefits — the huge impact over time of U.S. government spending in defense, disease control, and space to the private business and the economy (and, thus, in the current debate, too).
In a developed economy, the resulting finer accounting of the sources of growth can effectively inform policy making. In revealing more of the reality through the model, it ought to clarify the current debate and create a non-ideological basis for new policy.
Applications in Latecomer Countries. As an innovation practitioner from a latecomer country, I believe Romer’s research can also be used to design innovative products, institutions, and linkages. This really applies at the diffusion end, and not the invention end, of the innovation process where technology is most relevant to latecomer countries. In showing governments where to create winners (and not just intervene in case of market failures), it can also inform policy from the micro-level like in geographic and sector clusters.
Romer’s research appears to focus on technology. For latecomer countries, the use of a broader definition of technology from one based narrowly on science and technology to one to include its “systematic application” creates the opportunities. Here we need to go to the details.
Firstly, Romer saw technology as non-rival. It cannot be used up even when shared. If distributed at low cost, say through the Internet, the effective fixed cost per unit is reduced according to its available market and so results in increasing returns.
Secondly, he saw technology as partially excludable. That means one can appropriate ownership say by patents, branding, embedding in products, channel control, creative distribution, etc. This feature is important because it allows lock-in and recovery of values.
Monopolistic Competition is Norm in Technology Space. There is some concern that, because non-rival and partially excludable, monopolistic competition can occur in technology space.
Other research has shown that despite apparent imperfect competition, monopoly may not necessarily follow. Rapid technology obsolescence and convergence limits monopoly. Also, congestion in the technology channel as with bandwidth, speed of diffusion, physical limits like atomic sizes in the case of circuit widths, will likewise constrict monopoly power.
Innovative Applications in Romer Space. By looking for applications and product models in non-rival, partially excludable Romer space, I hope to help bring his basic research to the upper right quadrant of the typical innovation S-curve. I have found a few models:
- Embedded knowledge like vitamin-fortification in bread. Romer’s example is oral rehydration therapies. The distribution cost for this type is high because the carrier products are hard. The embedded knowledge in the fortification and therapy design belong to Romer space.
- Potentially viral products like Black-Eyed Peas “I Gotta Feeling” with the bundled flash mob dance steps. In this paired cliché – as in song and dance, the song is partially excludable and the dance is non-rival and free. This is interesting because the non-excludable part is not technology but entertainment; technology in this case is the enabler for distribution and to assure excludability.
- In Steve Job’s more complex product architecture, the bundles of iPod and iTunes and the iPhone and App Store also exist in Romer Space.
- Aplia, Romer’s own internet-enabled e-learning business, has a similar architecture as You Tube. The soft content is free or discounted versus the equivalent hard goods and partially excludable via channel control. They have similarly lumped fixed investment to set up the Internet business but the non-rivalry of the content guarantees profits in the long run.
- Mobile phone banking for microfinance. In the Philippines, this can be the super-enabler by bringing the credit multiplier via PayPal-like electronic wallets to the previously inaccessible countryside at non-rival costs. There are enabling institutions and linkages needed to make this work.
From these examples, infinite variants of new goods are possible in Romer space. Non-rival components can come from knowledge or experience. Partial excludability components can come from patents, branding, embedding in products, channel control, creative distribution. The limit is our creativity.
So from the basic research to the book to applications in the ground, the epic voyage continues.