Category Archives: cities

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.

Accessibility, Mobility and Density

Are accessibility and mobility complements or substitutes? I have a mental model a graph with a y-axis as density, and x-axis as mobility, where the Northeast corner would be high access: high density multiplied by high mobiilty.
This system behaves differently by modes. For transit, cities arrange themselves on a line from the southwest to the northeast (a positive feedback loop between supply and demand). For auto cities arrange on a line from the southeast to the northwest (a negative feedback loop between congestion and demand). Using data one could place specific cities on the graph. One expects places like New York and Hong Kong in the northeast corner, most US cities in the southeast corner, small developing-world cities without widespread adoption of modern automobile or transit technology in the southwest corner. Depending on where you draw the threshold, it is hard to see too many places in the upper northwest corner, as it would be difficult to grow to have high density without mobility. (Why would the city grow without the accessibility advantages?)
Density Mobility Tradeoff
Accessibility is a good, but it is not a good without costs, and there are limits to how much people are willing to pay for access. It may also suffer from diminishing returns, beyond a point each unit of accessibility is worth less and less. Places like Minneapolis have yet to reach that point, but surely there are places that have.

The Elements of Vibe

What is vibe? Vibe is the vitality of street life, the feeling that there is something going on, of being where the action is. Successful places have vibe, dead places don’t. We don’t want vibe everywhere and probably can’t support it. But surely we could have more active places then we do now with a better location of activities.
We drive to places we can walk around, rather than walk around our own neighborhood, unless we happen to live in a place with vibe.
Why do we want to walk around? Because there are multiple things to do: find food, browse books, hear music, entice the intellect, stimulate the senses. This concentration of activities only happens because of the crowds around, and the crowds only gather because of the concentration. More begets more.
These are ‘economies of agglomeration’ as the economists might say or perhaps ‘network effects’. But they allow for the spontaneous walk-in business rather than the planned trip. Many businesses are unlikely to attract spontaneous walk-ins, for instance vacuum cleaner repairs, [I don't normally walk around with a vacuum cleaner on the hope I will find a repair shop] and thus lose little by not being located in the center of action and save much on rent. Some restaurants are so good, they require a reservation, and thus there is little spill-in traffic. But other businesses, by saving on rent, are foregoing additional business.
Moreover, those businesses are denying potential spillover traffic to their would-be neighbors. It is a calculation that proprietors must do for themselves, but there is a coordination function that a good entrepreneur can serve, matching businesses that attract walk-ins with compatible stores, and maybe subsidizing (lowering the rent for) those that generate more spill-over traffic than they attract.
There are three seeds:
* A concentration of people (customers, though they need not be spending money, that helps)
* A concentration of stuff (suppliers, who need not be selling)
* An environment that encourages people to spend time doing stuff (marketplace)
People concentrate for a variety of reasons – to exploit the material resources of the earth, to have safety in numbers, to find a pool of potential mates, or simply because it is at the intersections of routes between two other places. These intersections (nodes in transportation lingo), create opportunities. In the streetcar era, people might change lines at a node, and those pedestrians would create the streetlife necessary to support new businesses. In the highway era the scale changed, and nodes are the interchanges of freeways. Businesses, and especially shopping malls, take advantage of these points of high accessibility. But the shopping mall is now clearly the destination, not a side-product of a transfer point in the same way street-car corners were.
Some further assertions about human nature:
People like pleasant climates – dry, not too hot, not too cold, clean air, not too loud.
People want to feel safe – they don’t want a car careening out of control disturbing their sidewalk café meal, they don’t want to think they will get run over crossing the street.
People are lazy – they don’t want to walk too far to get where they are going. If they are driving, they want easy convenient parking near their destination. They like to cross the street midblock and don’t want to have to walk to intersections.
People are cheap – they don’t want to pay for that easy convenient parking, they prefer lower to higher prices for the same good.
The last two be summarized by the idea that “People take the path of least resistance�?.
Observing cities around the world with an informed, but casual analysis leads me to assert some rules about the environment that lead to vibrancy.
Buildings on the sidewalk – vibrant areas have buildings that abut sidewalks with not large gaps between the building and the walk. The density of activity is necessarily reduced by space between building and path (and thus other buildings).
Sidewalks on the street – to have vibe, sidewalks must abut the street, or *be* he street in pedestrian only areas. Pedestrian only areas can work, and anyone who says otherwise has other interests at heart. This does not mean that they will work, but given the right environment, people would prefer to shop without having to look out for motorized vehicles.
Streets move slowly – fast streets make pedestrians feel unsafe, and thus reduces the benefits of being on the sidewalk. Ideally streets are moving at pedestrian speed in the pedestrian area. Of course streets leading to the pedestrian area move faster, or people could not get there.
Vehicle space on the street is minimal – wide streets increase the distance pedestrians must walk to reach other activities. Narrow streets give access to more stuff in less time. Hence the reason many enclosed shopping malls work better than many shopping streets is the density of stuff is fairly tight.
Street two way – One way streets may not be inherently problematic, but one-way streets are generally that way to move more vehicle traffic faster through the area, which is the opposite goal of moving pedestrians between buildings within the area.
Opportunities to explore just around the corner – hidden (pleasant) surprises are one of the things that make cities interesting to be in, if I go around this corner what will I discover. The same opportunities do not exist in an enclosed shopping mall, where everything is pre-mapped and tightly controlled, and I know each “block” ends at a parking ramp. Hidden unpleasant surprises however are one of the things that can kill a city, I don’t want to experience dread when I walk down an alley attached to my favorite shopping street.
This set of rules is by no means complete, but rules like these created streetlife in streetcar era places, and they create vibe in the better shopping malls. 


In praise of landmarks

Yesterday, Apple Computer announced a slew of new products, among them iTunes 7. A key feature of this piece of software is its new user interface, dubbed “CoverFlow”, discussed in this article: Wired News: New UI Showdown: Apple vs. TiVo. You do need to see it to fully understand it, and it is quite a slick way to navigate a music database.
Why am I talking about music databases and album art in a blog about transportation? I think cities are much like databases, and buildings like album covers. We navigate spatially and visually. Cities without redeeming art, architecture or natural landmarks are unpleasant. Not merely because they lack “charm” and the buildings are individually dull, but because of their collective undifferentiatedness, which creates difficulties for navigating (especially if they also lack some spatial regularity like a comprehensible grid network) and spatially locating oneself. Being lost (both not knowing where you are and not knowing either how you got there or how you will get to where you are going) brings a strong sense of unease that creates frustration if not hostility to the place you are lost at.
Cities need the equivalent of album art so that people can explore them. The nature of this art changes if you are walking, biking, taking transit, or driving, as you view it at different speeds and different resolutions.
Skylines have value, more than the simple value to the owner of the individual building. (In economic terms, they provide some positive externality, collectively exhibiting a network effect where the whole is larger than the sum of the individual parts. Measuring this is of course difficult.)
When I am driving around, and see the skyline in the distance from a particular angle, I instantly know what direction I am going. While some of these benefits may be obviated with in-vehicle navigation, the certainty of physical structure outweighs the digital outputs of a machine.
I recall as a freshman at Georgia Tech, taking a night course, leaving some classroom building for the first time out of a door different from where I entered and being completely turned around, until eventually I located the Coca-Cola headquarters building (just southwest of campus). While I still didn’t know exactly where I was, I could figure out where I was going.