The Iron Laws of Network Cost Scaling

Eric Raymond posits The Iron Laws of Network Cost Scaling (which he applies strictly to communications networks).
1. Upgrade cost per increment of capacity decreases as capacity rises.
2. Network costs scale primarily with the number of troubleshooters required to run them, not with capacity.
3. Under market pressure, network pricing evolves from metered to flat-rate.
I am not sure I buy these as laws, though I don’t necessarily disagree with as general observations,
1. is economies of scale of capital costs, which I think do operate in telecommunications in a way differently than transportation
2. is operations and maintenance, which is a function of demand as much as supply. Again telecom differs from transportation
3. assumes we are on a downward sloping average cost curve. Transportation professionals are trying to move pricing in the other direction (to date, largely unsuccessfully), especially in congested areas, where costs slope upwards.

xkcd: Long Light

From xkcd: Long Light
Linked to in comments in an article in Slashdot: “Traffic-Flow Algorithm can reduce fuel consumption

Economic Development Impacts of High Speed Rail

Recent working paper (extending recent blog posts):

This paper reviews the state of high-speed rail (HSR) planning in the United States c. 2010. The plans generally call for a set of barely inter-connected hub-and-spoke networks. The evidence from US transit systems shows that lines have two major impacts. There are positive accessibility benefits near stations, but there are negative nuisance effects along the lines themselves. High speed lines are unlikely to have local accessibility benefits separate from connecting local transit lines because there is little advantage for most people or businesses to locate near a line used infrequently (unlike public transit). However they may have more widespread metropolitan level effects. They will retain, and perhaps worse, have much higher, nuisance effects. If high-speed rail lines can create larger effective regions, that might affect the distribution of who wins and loses from such infrastructure. The magnitude of agglomeration economies is uncertain (and certainly location-specific), but presents the best case that can be made in favor of HSR in the US.