Bettencourt, L.M.A., J.Lobo, D. Helbing, C. Kühnert, and G.B. West (2007) Growth, innovation, scaling, and the pace of life in cities."

This paper was absolutely awesome. Almost right up my alley. They didn't just take lots of data and call it all scale free. They compared different types of urban data and compared the exponents, describing the implications of their very abstract and high level views for real policy implications. Intrigueingly, Lin Ostrom is the editor of this paper, which convinces me further that she is the person to talk to if I'm going to bridge complex systems and policy.

They gloss scaling exponents > 1 as signifying increasing returns and exponents<1 as economies of scale. Still have to get my head around that. I never though of those two as opposites.

They talk abou tthe fact the quarter scaling ubiquitous in biology isn't so relevant to cities, and why. They talk about its relationship to the metaphors. How cool that phrases like 'biological metaphor' and 'pace of life' are in PNAS.

Their central graph, showing how population, housing, roads, wires, crime and countless measures of innovation scale with city size is great. Exceptions to all the >1 results (cases when crime doesn't scale >1) could provide arguments for good policy and the nature of the exponents. It is interesting to read the graph as providing environmental arguments against local, small, rural low density living.

Rough dichotomy of infrastructure networks (<1) and social networks (>1)

"This difference can only be reconciled if the distribution netowrk is suboptimal, as observed in the scaling of resistive losses" I'd love to know what that means.

I'd need to think more about exponential growth of some phenom as the engine of emergence of higher scale order, and the mechanisms by which such growth eventually tops off as selecting for 'which' higher scale of order.

Their results challenge classic results in urban econ: Henderson JV 74, Henderson JV 88, Drennan MP 2002.

Neat: "New indices of urban rank according to deviation from the predictions of scaling laws also provide more accurate measures of the successes and failures of local factors (including policy) in shaping specific cities" That has implications...

All their data was from Census. US, EU and China.