Wednesday, February 13, 2013

Arab Spring or Decentralization


Joseph S. Nye Gets It …
 … in part at least. In a post for Project Syndicate (Feb. 7, 2013), he (re)discovers the analysis I developed in “The Second Twentieth Century” (2000 Grasset, 2006 Hoover Press): the information revolution started in the 1970s, the dramatic decrease in the cost of producing, storing and transmitting information, has a major influence on the decentralization of large hierarchies, including firms and governments. It is now underlying the “Arab Spring” that is really an Arab political revolution.

Excerpts:

“The current global revolution is based on rapid technological advances that have dramatically decreased the cost of creating, finding, and transmitting information.” This is my main thesis in the book, expressed in the very same terms.

“As computing power has become cheaper and computers have shrunk to the size of smart phones and other portable devices, the decentralizing effects have been dramatic.”  This is my conclusion and, as I show in the book, the result of the “Coase-Rybczinski theorem”: information-intensive organizations (decentralized, market using processes) develop and centralized ones regress when information gets cheaper.

“The spread of information means that informal networks are undercutting the monopoly of traditional bureaucracy”. Idem. So thank you for reading my book and using it (but an honest quote would have helped).

Nye however dismisses the extreme case of complete decentralization as unlikely to happen and takes it as a proof of the limitation of the analysis:

“While the information revolution could, in principle, reduce large states’ power and increase that of small states and non-state actors, politics and power are more complex than such technological determinism implies.”

That conclusion is trivial but the analysis is false because Nye does not understand basic microeconomic analysis and relies on the simplistic view of a trend-like evolution going to the extremes: it is true indeed in economics that we rarely get a “corner solution”, the extreme case in an all or nothing approach of a phenomenon. Thus we should not expect the information revolution to lead to a complete decentralization of all governments and of all large firms. There exists, somewhere, an optimal solution to the decentralization trend, in which some large hierarchies subsist side by side with smaller ones, including some single-entrepreneur (or artisanal) businesses. But nevertheless a continued fall of the cost of information will push all organizations towards a smaller-size equilibrium.

This has nothing to do with a contrast between a “principle” and “reality”, the latter meaning that “politics and power are more complex than such technological determinism implies” as Nye claims.

It is simply due to the basic economic price mechanism in which a “corner solution” (the extreme case) can only be obtained if the price of one of the objects of choice falls or rises to an extreme. In the present case complete decentralization would occur if the cost of information were to fall to zero, an unlikely event to say the least.

The conclusion is that the analysis of organizations and hierarchies, i.e. of “institutions”, is not a matter of economic “principle” versus political “realism”: it is amenable to sound economic analysis, correctly understood and applied, period.

No comments: