Algorithmic Economy

Hezekiah Agwara, Philip Auerswald, and Brian Higginbotham at George Mason University recently published an important new paper on “Algorithms and the Changing Frontier“. Given at an NBER conference on The Changing Frontier: Rethinking Science and Innovation Policy, the paper actually covers two topics: a discussion of the algorithmic economy and the role of standards in globalization (in the context of the algorithmic economy).
On the first topic, they lay out an interesting refinement of what a number of us have been discussing as the knowledge economy:

In the science-based (aka New Growth) model, technological distance does not exist; newly discovered recipes add to aggregate knowledge as soon as they are put into practice. In the algorithmic model, search for better recipes is constrained both by technological distance and by the complexity of the production process. Newly discovered recipes that are not easily imitated are the essence of economic differentiation and the basis for above-normal profits; the interoperability of recipes is essential to the functioning of complex supply chains.
In this light, consider the notion, central to the science-based model, that both ideas are “nonrival” and “non-excludable,” economically relevant innovations are characteristically subject to “knowledge spillovers.” In the algorithmic model, the ideas that actually propel growth and development are overwhelmingly uncodified, context dependent, and transferable only at significant cost–which is to say that tacit knowledge dominates, information asymmetries are the norm, and transactions costs are significant.
While knowledge spillovers of the type that are central to the science-based model clearly exist, they are unlikely to be of significant relevance in the practical work of creating the new business entities that drive twenty-first-century global value chains. The reason for this is that most productive knowledge is firm specific and producers far from dominant production clusters must learn to produce through a process of trial and error. Market-driven innovation involves the search for ideas that are rivalrous and excludable (at least temporarily), out of which ventures with proprietary value can be created. The impediments to innovation that matter most are not a lack of appropriability of returns but the everyday battles involved in communicating ideas, building trust, and making deals across geographically disparate regions and diverse economic units (Auerswald, 2008).
. . .
The emphasis of the science-based model on product innovation naturally leads to a view of the economic frontier in which technology adoption or transfer (largely based on technical standards) are the main conduits for global innovation and knowledge stocks are well represented by patents. The algorithmic model begins with the premise that product innovation is impossible without process innovation; the conversion of new or improved products, and the related technical standards, into commercial products of global value requires substantial innovation in production processes. Not all firms are at the boundaries of the production possibilities frontier, and differences in the quality of operational processes could separate firms with largely similar technical capabilities. Symmetric access to product innovations does not suggest a convergence in productivity.

In other words, development of a technology is not enough. While “technologies” (aka ideas/recipes) are nonrival and non-excludable, implementation can (and does) create excludability.
This insight goes beyond the process of technology development. I have always argued that knowledge economy is broader than just science-based. So what is interesting about this discussion is that the insight can (and should) be applied to all knowledge-based innovation — not just science-based. Substitute “business model” for “technology” and “product development” in the above discussion, and ideas remain the same. The key to success in the “knowledge” economy is not simply the creation of knowledge but the ability to utilize that knowledge.
Thus, I would argue that innovation requires a host of other concomitant investments that go well beyond the standard Washington S&T consensus — i.e. fund R&D and innovation will take care of itself. We need a revamped innovation policy that understands this point. We need a new consensus that innovation, not just science, is the real endless frontier.
Just a quick word on the second topic of standards and their place in globalization. The authors find a correlation between acceptance of ISO 9000 and various knowledge-economy measures. They describe three reasons:

1)”ISO certification eases entry into production and distribution networks on a global scale because it signals a willingness and a capability for low-transactions cost integration into global production networks.”
2) “. . . the process of completing ISO certification serves as a learning tool that materially improves productive efficiency . . .”
3) “ISO certification increases functional compatibility and interoperability according to global norms, and thus eases adoption of platform technologies, or general-purpose technologies.”

Thus standards are an important part of spreading the algorithmic economy.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s