Paul Romer's approach to acknowledging the need for "rule" evolution and competition for best practice is very compelling. Indeed it is the basis for a Federalist system. One could argue that the ongoing centralization of US policy via federal regulation has hindered the countries rate of adaptation. The desire for regulatory unity and standards is helpful, but if it oversteps can lead to ossification. SEZ (special economic zones) and charter cities as put forth by Romer are needed now more than ever. Our technology and rapid rate of idea inter-exchange mean that we need our rules to evolve at an ever faster rate.
The implicit question posed by this argument: "is the US becoming a dated mainframe type structured system in an ever evolving eco-structure of laws and economics that needs smaller units of response at state and local levels." The technologies of biotechnology alone are going to change social stresses faster than computers in the next 30 years. How will we evolve our rules, be they codified laws, social mores or behaviour sets? Should the federal level be locking down a single rule set or should state and local systems be given more freedoms? Premature optimization is a problem in computer coding and in "rule establishment" which has very real costs, think of the federal prohibition experiment in the 20's. The flipside of this argument is of course the Jim Crow laws of the south.
I like Romer as he reminds me of Hernando de Soto (Video here) recognizing the role that behaviours codified as laws or informally as acknowledged as cultural norms drive the human economic and social condition. He gets the meta game, even if he acknowledge that the Jeffersonian Federalist concept was there before. The historical case of Hong Kong put forth by Romer is a bit stretched in the discussion below and should be taken with a grain of salt relative to its very awkward origins as a drug runners paradise run by the British after the seccond opium wars.
