How constraints drive growth

A city is a positive feedback loop in space.
Why locate anywhere but to be near something or far from something? Cities offer opportunities to be near lots of things (people), and as cities exist, those things must be of value to the people who locate there. By locating, people add to the “stuff” others can reach.
Accessibility is a measure of nearness to things (people), e.g. how much stuff you can reach in x minutes time. Which stuff matters and how much time is acceptable depend on individual preferences, but these can be measured and observed. An area with higher density enables you to reach more stuff in less time because it is physically closer, even if the network is slower (you can move less distance per unit time), provided the density increases at a rate faster than slowness increases.
Some cities are physically constrained, notably San Francisco (a peninsula) and Manhattan and most of New York (islands). In fact, the five densest cities in the US (New York, San Francisco, Los Angeles, Honolulu, and Chicago) all have some significant physical constraints (island, peninsula/bay, mountains/ocean, island/mountains, lake) hemming them in. Not surprisingly, these are among the most expensive cities in which to live. This indicates that the location is especially valuable, because of the accessibility benefits it provides.
Perhaps it is the constraint itself which creates value. Because of the constraint, more people and firms are bidding for scarce space (since the non-scarce non-central available space has a much higher transportation cost (across the bay, off the island, in more distant suburbs) driving up rents. As a consequence, developers build at higher density in the core city, increasing accessibility. Because of the higher density, there is higher accessibility, creating value for residents and businesses, leading to even higher rents. Location has positive spillovers.
A city is a positive feedback loop in space. Spatial constraints accelerate the loop.

How volatility begets stability

The human body evolved over time through natural selection. Until very recent advances in medicine which have yet to affect evolution, the entire body depends on various components (heart, lung, brains, etc.), and if any one of them fails, the whole fails. Thus it made no sense to evolve a brain that would noticeably outlast the heart, or a heart that would outlast the lungs. Any effort in a longer-lived brain would be moot as the heart would fail first, and similarly, any attempt to have a heart that would beat longer than the lungs could breathe would be over-engineering. The marginal rate of return of extending the life of any critical organ would probably be equalized in such a scenario.
A similar logic has been alleged to apply to autos, with planned obsolescence. Why design a frame that outlasts the engine? The ideal, from a narrow efficiency point of view, is for all parts to fail simultaneously with no point in spending money on repairs, and no excess wasted at the outset by having parts last longer than the whole.
Both the human body and technological artifacts like automobiles are finite systems. While the date of reckoning for bodies or cars may not be known in advance, nobody naturally lives past about 120 and intensively used cars do not economically (or typically) last past 20 years.
But there are other systems that are potentially infinite. These include cities and networks. While a city or network may not last forever, its potential lifespan is quite uncertain (we may assume that if the sun dies out and planet earth is abandoned, cities will be too, but the 5,000,000,000 year upper limit is beyond comprehension). These potentially infinite systems last significantly longer than their component artifacts. Just as Heraclitus said ‘no one ever steps into the same river twice’, one never steps into the same city twice, it is continuously evolving, as parts are abandoned, destroyed, replaced, or rebuilt. Quite often, the city while changing its buildings, maintains its networks, whose topographical (and topological) structure outlasts its buildings, in part due to property ownership regulations. (The example of London being rebuilt using essentially the same routes after the 1666 fire being the most illustrative of this case).
I argue the reason that cities (and their networks) last longer is that their components fail at different rates. If all of the components (buildings, plexus (networks, social structures)) failed at the same time (e.g. a fire plus a breakdown of the legal system ensuring property rights), then the site could be abandoned. But as long as most components last, a few failing will not destroy the city. The resources from the remaining components can help rebuild the failed ones. Similarly, networks do not fail together, and the failure of one link (given some redundancy) will not cause the network to collapse.
This volatility in failure rates of components leads to a more stable whole. The price is that only piecemeal, rather than systematic overhaul of the system is permitted.

Problems at Elsevier

From the Ouroboros blog: Elsevier publishes fake journal for Merck
(Elsevier publishes the well-known Transportation Research parts A – F and other journals in the transportation field (Research in Transportation Economics, Transport Geography, Accident Analysis and Prevention, Transport Policy, and Journal of Air Transport Management) (disclosure: in which journals I have published and/or still have papers under review, but the long review process is another complaint)).
Read the article and the links within the article (especially The Tree of Life) for some fascinating information about the state of for-profit academic publishing.
Ouroboros writes

When it’s not gouging academic libraries with outlandish subscriptional fees, Elsevier finds other ways to boost its bottom line: Publishing bogus journals for pharmaceutical companies. From The Scientist:

Merck paid an undisclosed sum to Elsevier to produce several volumes of a publication that had the look of a peer-reviewed medical journal, but contained only reprinted or summarized articles–most of which presented data favorable to Merck products–that appeared to act solely as marketing tools with no disclosure of company sponsorship.


Another critique of Elsevier’s practices from Ars Technica at:
Editorial: publishing economics harm science’s credibility

China bridge collapse

From The Financial: Six killed in central China bridge collapse

Six killed in central China bridge collapse
18/05/2009 10:41 – (07:13 minutes ago)
The FINANCIAL — BEIJING, Six people have died after a road bridge collapsed in central China’s Hunan Province, Xinhua said on May 18.
The bridge, which was closed by government on May 5 to be dismantled, fell on a busy road in downtown Zhuzhou city on May 17, burying 26 cars and a bus.
Another 17 people were injured and a rescue operation is underway.
Workers said the collapsed section of the bridge will be exploded on May 20.

Better Place EV system

From: International Business Times:
Better Place unveils tech to switch EVs batteries
Better Place is a company building a network of charging stations, and more interestingly, battery replacement stations for upcoming standardized EVs using lithium ion batteries. This is the kind of scheme that has a very high fixed cost for deployment, but potentially low variable costs downstream, and only works if it has a lot of EVs over which to spread the initial cost of deployment, and therefore needs patient capital. The whole scheme, led by Shai Agassi a software entrepreneur, and plans to deploy initially in Israel, which has obvious reasons for wanting to switch from Oil to electric.
A video of the battery replacement (say for a long trip that would deplete the standardized battery technology they propose where it is faster to replace than recharge) is below